• Themes
• Newsmakers

# Fusion

17.08.2021
18:40 Laser Fusion Experiment Unleashes an Energetic Burst of Optimism

Even scientists who were skeptical of work at the National Ignition Facility called the results a success.

15:53 Strong magnetic fields change how friction works in plasma

Friction in plasma gets weird in the presence of very strong magnetic fields, a team of plasma researchers at the University of Michigan has shown. The findings could affect fusion energy strategies and the development of radiation sources.

06:41 Accelerating the estimation of energetic particle confinement statistics in stellarators using multifidelity Monte Carlo. (arXiv:2108.06408v1 [physics.plasm-ph])

In the design of stellarators, energetic particle confinement is a critical point of concern which remains challenging to study from a numerical point of view. Standard Monte Carlo analyses are highly expensive because a large number of particle trajectories need to be integrated over long time scales, and small time steps must be taken to accurately capture the features of the wide variety of trajectories. Even when they are based on guiding center trajectories, as opposed to full-orbit trajectories, these standard Monte Carlo studies are too expensive to be included in most stellarator optimization codes. We present the first multifidelity Monte Carlo scheme for accelerating the estimation of energetic particle confinement in stellarators. Our approach relies on a two-level hierarchy, in which a guiding center model serves as the high-fidelity model, and a data-driven linear interpolant is leveraged as the low-fidelity surrogate model. We apply multifidelity Monte Carlo to the study

05:48 Accelerating the estimation of energetic particle confinement statistics in stellarators using multifidelity Monte Carlo. (arXiv:2108.06408v1 [physics.plasm-ph])

In the design of stellarators, energetic particle confinement is a critical point of concern which remains challenging to study from a numerical point of view. Standard Monte Carlo analyses are highly expensive because a large number of particle trajectories need to be integrated over long time scales, and small time steps must be taken to accurately capture the features of the wide variety of trajectories. Even when they are based on guiding center trajectories, as opposed to full-orbit trajectories, these standard Monte Carlo studies are too expensive to be included in most stellarator optimization codes. We present the first multifidelity Monte Carlo scheme for accelerating the estimation of energetic particle confinement in stellarators. Our approach relies on a two-level hierarchy, in which a guiding center model serves as the high-fidelity model, and a data-driven linear interpolant is leveraged as the low-fidelity surrogate model. We apply multifidelity Monte Carlo to the study

05:47 Accelerating the estimation of energetic particle confinement statistics in stellarators using multifidelity Monte Carlo. (arXiv:2108.06408v1 [physics.plasm-ph])

In the design of stellarators, energetic particle confinement is a critical point of concern which remains challenging to study from a numerical point of view. Standard Monte Carlo analyses are highly expensive because a large number of particle trajectories need to be integrated over long time scales, and small time steps must be taken to accurately capture the features of the wide variety of trajectories. Even when they are based on guiding center trajectories, as opposed to full-orbit trajectories, these standard Monte Carlo studies are too expensive to be included in most stellarator optimization codes. We present the first multifidelity Monte Carlo scheme for accelerating the estimation of energetic particle confinement in stellarators. Our approach relies on a two-level hierarchy, in which a guiding center model serves as the high-fidelity model, and a data-driven linear interpolant is leveraged as the low-fidelity surrogate model. We apply multifidelity Monte Carlo to the study

16.08.2021
14:48 Can start-ups fast-track fusion energy?

08:27 Heat conduction in an irregular magnetic field: Part II. Heat transport as a measure of the effective non-integrable volume. (arXiv:2108.06328v1 [physics.plasm-ph])

Given the large anisotropy of transport processes in magnetized plasmas, the magnetic field structure can strongly impact the heat diffusion: magnetic surfaces and cantori form barriers to transport while chaotic layers and island structures can degrade confinement. When a small but finite amount of perpendicular diffusion is included, the structure of the magnetic field becomes less important, allowing finite pressure gradients to be supported across chaotic regions and island chains. We introduce a metric for the effective volume of non-integrability based on the solution to the anisotropic heat diffusion equation. To validate this metric, we consider model fields with a single island chain and a strongly chaotic layer for which analytic predictions of the relative parallel and perpendicular transport can be made. We also analyze critically chaotic fields produced from different sets of perturbations, highlighting the impact of the mode number spectrum on the heat transport. We

12.08.2021
21:44 The Wendelstein 7-X concept proves its efficiency

One of the most important optimization goals underlying the Wendelstein 7-X fusion device has now been confirmed. An analysis shows: In the optimized magnetic field cage, the energy losses of the plasma are reduced in the desired way. Wendelstein 7-X is intended to prove that the disadvantages of earlier stellarators can be overcome and that stellarator-type devices are suitable for power plants.

19:52 The Wendelstein 7-X concept proves its efficiency

One of the most important optimisation goals underlying the Wendelstein 7-X fusion device at Max Planck Institute for Plasma Physics (IPP) in Greifswald has now been confirmed. An analysis by IPP scientists in the journal Nature shows: In the optimized magnetic field cage, the energy losses of the plasma are reduced in the desired way. Wendelstein 7-X is intended to prove that the disadvantages of earlier stellarators can be overcome and that stellarator-type devices are suitable for power plants.

06:24 Representation Learning for Remote Sensing: An Unsupervised Sensor Fusion Approach. (arXiv:2108.05094v1 [cs.CV])

In the application of machine learning to remote sensing, labeled data is often scarce or expensive, which impedes the training of powerful models like deep convolutional neural networks. Although unlabeled data is abundant, recent self-supervised learning approaches are ill-suited to the remote sensing domain. In addition, most remote sensing applications currently use only a small subset of the multi-sensor, multi-channel information available, motivating the need for fused multi-sensor representations. We propose a new self-supervised training objective, Contrastive Sensor Fusion, which exploits coterminous data from multiple sources to learn useful representations of every possible combination of those sources. This method uses information common across multiple sensors and bands by training a single model to produce a representation that remains similar when any subset of its input channels is used. Using a dataset of 47 million unlabeled coterminous image triplets, we train an

06:24 Towards Top-Down Just Noticeable Difference Estimation of Natural Images. (arXiv:2108.05058v1 [cs.CV])

Existing efforts on Just noticeable difference (JND) estimation mainly dedicate to modeling the visibility masking effects of different factors in spatial and frequency domains, and then fusing them into an overall JND estimate. However, the overall visibility masking effect can be related with more contributing factors beyond those have been considered in the literature and it is also insufficiently accurate to formulate the masking effect even for an individual factor. Moreover, the potential interactions among different masking effects are also difficult to be characterized with a simple fusion model. In this work, we turn to a dramatically different way to address these problems with a top-down design philosophy. Instead of formulating and fusing multiple masking effects in a bottom-up way, the proposed JND estimation model directly generates a critical perceptual lossless (CPL) image from a top-down perspective and calculates the difference map between the original image and the

06:24 Learning Deep Multimodal Feature Representation with Asymmetric Multi-layer Fusion. (arXiv:2108.05009v1 [cs.CV])

We propose a compact and effective framework to fuse multimodal features at multiple layers in a single network. The framework consists of two innovative fusion schemes. Firstly, unlike existing multimodal methods that necessitate individual encoders for different modalities, we verify that multimodal features can be learnt within a shared single network by merely maintaining modality-specific batch normalization layers in the encoder, which also enables implicit fusion via joint feature representation learning. Secondly, we propose a bidirectional multi-layer fusion scheme, where multimodal features can be exploited progressively. To take advantage of such scheme, we introduce two asymmetric fusion operations including channel shuffle and pixel shift, which learn different fused features with respect to different fusion directions. These two operations are parameter-free and strengthen the multimodal feature interactions across channels as well as enhance the spatial feature

11.08.2021
18:37 Demonstration of reduced neoclassical energy transport in Wendelstein 7-X

09:02 TBNet:Two-Stream Boundary-aware Network for Generic Image Manipulation Localization. (arXiv:2108.04508v1 [cs.CV])

Finding tampered regions in images is a hot research topic in machine learning and computer vision. Although many image manipulation location algorithms have been proposed, most of them only focus on the RGB images with different color spaces, and the frequency information that contains the potential tampering clues is often ignored. In this work, a novel end-to-end two-stream boundary-aware network (abbreviated as TBNet) is proposed for generic image manipulation localization in which the RGB stream, the frequency stream, and the boundary artifact location are explored in a unified framework. Specifically, we first design an adaptive frequency selection module (AFS) to adaptively select the appropriate frequency to mine inconsistent statistics and eliminate the interference of redundant statistics. Then, an adaptive cross-attention fusion module (ACF) is proposed to adaptively fuse the RGB feature and the frequency feature. Finally, the boundary artifact location network (BAL) is

10.08.2021
23:43 Scientists detect characteristics of the birth of a major challenge to harvesting fusion energy on Earth

Novel camera detects the birth of high-energy runaway electrons, which may lead to determining how to prevent damage caused by the highly energetic particles.

04:37 A Machine learning approach for rapid disaster response based on multi-modal data. The case of housing & shelter needs. (arXiv:2108.00887v2 [cs.LG] UPDATED)

Along with climate change, more frequent extreme events, such as flooding and tropical cyclones, threaten the livelihoods and wellbeing of poor and vulnerable populations. One of the most immediate needs of people affected by a disaster is finding shelter. While the proliferation of data on disasters is already helping to save lives, identifying damages in buildings, assessing shelter needs, and finding appropriate places to establish emergency shelters or settlements require a wide range of data to be combined rapidly. To address this gap and make a headway in comprehensive assessments, this paper proposes a machine learning workflow that aims to fuse and rapidly analyse multimodal data. This workflow is built around open and online data to ensure scalability and broad accessibility. Based on a database of 19 characteristics for more than 200 disasters worldwide, a fusion approach at the decision level was used. This technique allows the collected multimodal data to share a common

09.08.2021
06:29 Effect of soft and hard x-rays on shock propagation, preheating and ablation characteristics in pure and doped Be ablators. (arXiv:2108.02933v1 [physics.plasm-ph])

In this paper, we analyze the performance of pure and doped Be ablators used for Inertial Confinement Fusion (ICF) pellets in terms of shock velocity, shock breakout temperature, preheat temperature and mass ablation rate through radiation hydrodynamic (RHD) simulations. For this study, we apply a constant radiation profile (drive temperatures varying from 120 - 200 eV) consisting of a low frequency Planck spectrum (soft x-rays) and a high frequency Gaussian spectrum (hard x-rays, commonly termed as "M-band") on a planar foil of the ablator. The fraction of energy density in the hard x-ray spectrum ($\alpha$) has been varied from 0 to 0.25. The predominant effect of hard x-rays is to preheat the ablator ahead of the shock front. Steady rise in preheat temperature and shock breakout temperature is observed on increasing the fraction of hard x-rays. Preheating can be mitigated by doping Be with a mid-Z element Cu as its opacity is much higher in the high frequency region. On doping Be

05.08.2021
19:44 A major challenge to harvesting fusion energy on Earth

A key challenge for scientists striving to produce on Earth the fusion energy that powers the sun and stars is preventing what are called runaway electrons, particles unleashed in disrupted fusion experiments that can bore holes in tokamaks, the doughnut-shaped machines that house the experiments. Scientists led by researchers at the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL) have used a novel diagnostic with wide-ranging capabilities to detect the birth, and the linear and exponential growth phases of high-energy runaway electrons, which may allow researchers to determine how to prevent the electrons' damage.

10:34 Pervasive Hand Gesture Recognition for Smartphones using Non-audible Sound and Deep Learning. (arXiv:2108.02148v1 [cs.SD])

Due to the mass advancement in ubiquitous technologies nowadays, new pervasive methods have come into the practice to provide new innovative features and stimulate the research on new human-computer interactions. This paper presents a hand gesture recognition method that utilizes the smartphone's built-in speakers and microphones. The proposed system emits an ultrasonic sonar-based signal (inaudible sound) from the smartphone's stereo speakers, which is then received by the smartphone's microphone and processed via a Convolutional Neural Network (CNN) for Hand Gesture Recognition. Data augmentation techniques are proposed to improve the detection accuracy and three dual-channel input fusion methods are compared. The first method merges the dual-channel audio as a single input spectrogram image. The second method adopts early fusion by concatenating the dual-channel spectrograms. The third method adopts late fusion by having two convectional input branches processing each of the

10:23 Exploration of the solar system and beyond using a thermonuclear fusion drive. (arXiv:2108.01689v1 [physics.pop-ph])

It is demonstrated that the development of a nuclear fusion rocket engine based on a D $-$ $^{3}$He (Deterium-Helium 3) technology will allow to travel in the solar system and beyond. The Direct Fusion Drive (DFD) is the D $-$ $^{3}$He-fueled, aneutronic, thermonuclear fusion propulsion system that is under development at Princeton University Plasma Physics Laboratory [1]. It is considered and analyzed the Earth-Mars mission using the DFD. It is shown that one-way trips to Mars in slightly more than 100 days become possible and also journeys to the asteroid belt will take about 250 days [2]. It is presented an analysis of realistic new trajectories for a robotic mission to Saturn's largest moon, Titan, to demonstrate the great advantages related to the thermonuclear DFD. The trajectories calculations and analysis for Saturn's largest moon Titan different profile missions are given based on the characteristics of a 2 MW class DFD engine. This capability results in a total trip duration

04.08.2021
06:23 More but Correct: Generating Diversified and Entity-revised Medical Response. (arXiv:2108.01266v1 [cs.CL])

Medical Dialogue Generation (MDG) is intended to build a medical dialogue system for intelligent consultation, which can communicate with patients in real-time, thereby improving the efficiency of clinical diagnosis with broad application prospects. This paper presents our proposed framework for the Chinese MDG organized by the 2021 China conference on knowledge graph and semantic computing (CCKS) competition, which requires generating context-consistent and medically meaningful responses conditioned on the dialogue history. In our framework, we propose a pipeline system composed of entity prediction and entity-aware dialogue generation, by adding predicted entities to the dialogue model with a fusion mechanism, thereby utilizing information from different sources. At the decoding stage, we propose a new decoding mechanism named Entity-revised Diverse Beam Search (EDBS) to improve entity correctness and promote the length and quality of the final response. The proposed method wins both

02.08.2021
04:27 Axisymmetric Plasma Equilibria with Toroidal and Poloidal Velocity Fields: Tokamak Relevant Configurations. (arXiv:2107.14766v1 [physics.plasm-ph])

We analyze an axisymmetric equilibrium of a plasma endowed with toroidal and poloidal velocity fields, with the aim to characterize the influence of the global motion on the morphology of the magnetic confinement. We construct our configuration assuming that the poloidal velocity field is aligned with the poloidal magnetic field lines and, furthermore, we require that the plasma mass density depend on the magnetic flux function (or equivalently, that the plasma fluid be incompressible). We then derive a sort of Grad-Shafranov equation for such an equilibrium and implement it to tokamak relevant situations, with particular reference to TCV-like profiles. The main result of the present study concerns the emergence, in configurations associated to a double-null profile, of a closed surface of null pressure encorporating the two X-points of the magnetic configuration. This scenario suggests the possible existence of a new regime of the plasma equilibrium, corresponding to an improved

04:27 Development of Compact Multivariable Probe for Two-Phase Detection in PbLi-Argon Columns. (arXiv:2107.14731v1 [physics.ins-det])

Liquid gas two phase flow is a common occurrence in various industrial applications. For nuclear fusion applications with a Li based breeder, existence of two phase flow may lead to critical issues including decreased tritium breeding ratio, generation of hot spots and improper nuclear shielding. Additionally, a very large density ratio of liquid metal to gas mandates relevant experiments towards development and validation of software tools. PbLi has gained immense focus for its various advantages and is utilized in several breeding blanket concepts. In view of above mentioned requirements, a two phase detection tool is imperative for liquid PbLi environment. For liquid metal applications, electrical conductivity based probes are most suitable in terms of ruggedness, fabrication ease and operational simplicity. However, corrosive nature and high operational temperature for PbLi put severe demands on electrical insulation, a foremost requirement for electrical conductivity based

30.07.2021
06:05 Multimodal Co-learning: Challenges, Applications with Datasets, Recent Advances and Future Directions. (arXiv:2107.13782v1 [cs.LG])

Multimodal deep learning systems which employ multiple modalities like text, image, audio, video, etc., are showing better performance in comparison with individual modalities (i.e., unimodal) systems. Multimodal machine learning involves multiple aspects: representation, translation, alignment, fusion, and co-learning. In the current state of multimodal machine learning, the assumptions are that all modalities are present, aligned, and noiseless during training and testing time. However, in real-world tasks, typically, it is observed that one or more modalities are missing, noisy, lacking annotated data, have unreliable labels, and are scarce in training or testing and or both. This challenge is addressed by a learning paradigm called multimodal co-learning. The modeling of a (resource-poor) modality is aided by exploiting knowledge from another (resource-rich) modality using transfer of knowledge between modalities, including their representations and predictive models. Co-learning

29.07.2021
08:50 Multi-Graph Convolutional-Recurrent Neural Network (MGC-RNN) for Short-Term Forecasting of Transit Passenger Flow. (arXiv:2107.13226v1 [cs.LG])

Short-term forecasting of passenger flow is critical for transit management and crowd regulation. Spatial dependencies, temporal dependencies, inter-station correlations driven by other latent factors, and exogenous factors bring challenges to the short-term forecasts of passenger flow of urban rail transit networks. An innovative deep learning approach, Multi-Graph Convolutional-Recurrent Neural Network (MGC-RNN) is proposed to forecast passenger flow in urban rail transit systems to incorporate these complex factors. We propose to use multiple graphs to encode the spatial and other heterogenous inter-station correlations. The temporal dynamics of the inter-station correlations are also modeled via the proposed multi-graph convolutional-recurrent neural network structure. Inflow and outflow of all stations can be collectively predicted with multiple time steps ahead via a sequence to sequence(seq2seq) architecture. The proposed method is applied to the short-term forecasts of

08:50 Squeeze-Excitation Convolutional Recurrent Neural Networks for Audio-Visual Scene Classification. (arXiv:2107.13180v1 [cs.MM])

The use of multiple and semantically correlated sources can provide complementary information to each other that may not be evident when working with individual modalities on their own. In this context, multi-modal models can help producing more accurate and robust predictions in machine learning tasks where audio-visual data is available. This paper presents a multi-modal model for automatic scene classification that exploits simultaneously auditory and visual information. The proposed approach makes use of two separate networks which are respectively trained in isolation on audio and visual data, so that each network specializes in a given modality. The visual subnetwork is a pre-trained VGG16 model followed by a bidiretional recurrent layer, while the residual audio subnetwork is based on stacked squeeze-excitation convolutional blocks trained from scratch. After training each subnetwork, the fusion of information from the audio and visual streams is performed at two different

08:06 Experimental Inference of Neutral and Impurity Transport in Alcator C-Mod Using High-Resolution X-Ray and Ultra-Violet Spectra. (arXiv:2107.13471v1 [physics.plasm-ph])

We present experimental inferences of cross-field impurity transport coefficients for Alcator C-Mod plasmas using a novel forward model for the entire Ca K-alpha spectrum, including satellite lines within the spectral range, to compare to high-resolution X-ray Imaging Crystal Spectroscopy (XICS). These measurements are complemented by Extreme Ultra-Violet (EUV) spectroscopy that constrains transport closer to the edge. Using new atomic data sets for both XICS and EUV analysis has enabled consideration of line ratios across both spectral ranges and has increased the accuracy of inferred transport coefficients. Inclusion of charge exchange between edge thermal neutrals and impurities is shown to be extremely important in C-Mod pedestals. We obtain D atomic neutral densities from experimental D Ly-alpha measurements at the midplane and compare these to SOLPS-ITER simulations, finding good agreement. Bayesian inferences of impurity transport coefficients are presented for L-, EDA H-, and

27.07.2021
06:31 MHD analysis on the physical designs of CFETR and HFRC. (arXiv:2107.11742v1 [physics.plasm-ph])

The China Fusion Engineering Test Reactor (CFETR) and the Huazhong Field Reversed Configuration (HFRC), currently both under intensive physical and engineering designs in China, are the two major projects representative of the low-density steady-state and high-density pulsed pathways to fusion. One of the primary tasks of the physics designs for both CFETR and HFRC is the assessment and analysis of the magnetohydrodynamic (MHD) stability of the proposed design schemes. Comprehensive efforts on the assessment of MHD stability of CFETR and HFRC baseline scenarios have led to preliminary progresses that may further benefit engineering designs.

23.07.2021
00:51 Gaming graphics card allows faster, more precise control of fusion energy experiments

Researchers have developed a method that uses a gaming graphics card to control plasma formation in their prototype fusion reactor.

22.07.2021
19:57 Gaming graphics card allows faster, more precise control of fusion energy experiments

Nuclear fusion offers the potential for a safe, clean and abundant energy source.

06:38 ECG Heartbeat Classification Using Multimodal Fusion. (arXiv:2107.09869v1 [cs.LG])

Electrocardiogram (ECG) is an authoritative source to diagnose and counter critical cardiovascular syndromes such as arrhythmia and myocardial infarction (MI). Current machine learning techniques either depend on manually extracted features or large and complex deep learning networks which merely utilize the 1D ECG signal directly. Since intelligent multimodal fusion can perform at the stateof-the-art level with an efficient deep network, therefore, in this paper, we propose two computationally efficient multimodal fusion frameworks for ECG heart beat classification called Multimodal Image Fusion (MIF) and Multimodal Feature Fusion (MFF). At the input of these frameworks, we convert the raw ECG data into three different images using Gramian Angular Field (GAF), Recurrence Plot (RP) and Markov Transition Field (MTF). In MIF, we first perform image fusion by combining three imaging modalities to create a single image modality which serves as input to the Convolutional Neural Network

06:04 Powder reuse cycles in electron beam powder bed fusion : Variation of powder characteristics. (arXiv:2107.09845v1 [cond-mat.mtrl-sci])

A path to lowering the economic barrier associated with the high cost of metal additively manufactured components is to reduce the waste via powder reuse (powder cycled back into the process) and recycling (powder chemically, physically, or thermally processed to recover the original properties) strategies. In electron beam powder bed fusion, there is a possibility of reusing 95 - 98% of the powder that is not melted. However, there is a lack of systematic studies focusing on quantifying the variation of powder properties induced by number of reuse cycles. This work compares the influence of multiple reuse cycles, as well as powder blends created from reused powder, on various powder characteristics such as the morphology, size distribution, flow properties, packing properties and chemical composition (oxygen and nitrogen content). It was found that there is an increase in measured response in powder size distribution, tapped density, Hausner ratio, Carr index, basic flow energy and

06:04 Turbulent field fluctuations in gyrokinetic and fluid plasmas. (arXiv:2107.09744v1 [physics.plasm-ph])

A key uncertainty in the design and development of magnetic confinement fusion energy reactors is predicting edge plasma turbulence. An essential step in overcoming this uncertainty is the validation in accuracy of reduced turbulent transport models. Drift-reduced Braginskii two-fluid theory is one such set of reduced equations that has for decades simulated boundary plasmas in experiment, but significant questions exist regarding its predictive ability. To this end, using a novel physics-informed deep learning framework, we demonstrate the first ever direct quantitative comparisons of turbulent field fluctuations between electrostatic two-fluid theory and electromagnetic gyrokinetic modelling with good overall agreement found in magnetized helical plasmas at low normalized pressure. This framework is readily adaptable to experimental and astrophysical environments, and presents a new technique for the numerical validation and discovery of reduced global plasma turbulence models.

20.07.2021
17:30 3D hohlraum model assists in indirect-drive implosions at NIF

Scientists from Lawrence Livermore National Laboratory (LLNL) and the Laboratory for Laser Energetics (LLE) have described a simple 3D model in hohlraums and capsules for inertial confinement fusion (ICF) implosions. The model will assist in delivering the required implosion symmetry on layered deuterium-tritium (DT) implosions for ignition.

04:37 Nucleation of helium in liquid lithium. (arXiv:2107.09017v1 [physics.class-ph])

Fusion energy stands out as a promising alternative for a future decarbonised energy system. To be sustainable, future fusion nuclear reactors will have to produce their own tritium. In the so-called breeding blanket of a reactor, the neutron bombardment of lithium will produce the desired tritium, but also helium, which can trigger nucleation mechanisms owing to the very low solubility of helium in liquid metals. An understanding of the underlying microscopic processes is important for improving the efficiency, sustainability and reliability of the fusion energy conversion process. A spontaneous creation of helium drops or bubbles in the liquid metal used as breeding material in some designs may be a serious issue for the performance of the breeding blankets. This phenomenon has yet to be fully studied and understood. This work aims to provide some insight on the behavior of lithium and helium mixtures at experimentally corresponding operating conditions (843 K and pressures between

16.07.2021
08:55 An Efficient and Small Convolutional Neural Network for Pest Recognition -- ExquisiteNet. (arXiv:2107.07167v1 [cs.CV])

Nowadays, due to the rapid population expansion, food shortage has become a critical issue. In order to stabilizing the food source production, preventing crops from being attacked by pests is very important. In generally, farmers use pesticides to kill pests, however, improperly using pesticides will also kill some insects which is beneficial to crops, such as bees. If the number of bees is too few, the supplement of food in the world will be in short. Besides, excessive pesticides will seriously pollute the environment. Accordingly, farmers need a machine which can automatically recognize the pests. Recently, deep learning is popular because its effectiveness in the field of image classification. In this paper, we propose a small and efficient model called ExquisiteNet to complete the task of recognizing the pests and we expect to apply our model on mobile devices. ExquisiteNet mainly consists of two blocks. One is double fusion with squeeze-and-excitation-bottleneck block (DFSEB

08:55 From Show to Tell: A Survey on Image Captioning. (arXiv:2107.06912v1 [cs.CV])

Connecting Vision and Language plays an essential role in Generative Intelligence. For this reason, in the last few years, a large research effort has been devoted to image captioning, i.e. the task of describing images with syntactically and semantically meaningful sentences. Starting from 2015 the task has generally been addressed with pipelines composed of a visual encoding step and a language model for text generation. During these years, both components have evolved considerably through the exploitation of object regions, attributes, and relationships and the introduction of multi-modal connections, fully-attentive approaches, and BERT-like early-fusion strategies. However, regardless of the impressive results obtained, research in image captioning has not reached a conclusive answer yet. This work aims at providing a comprehensive overview and categorization of image captioning approaches, from visual encoding and text generation to training strategies, used datasets, and

08:33 MuSIC@Indiana: an effective tool for accurate measurement of fusion with low-intensity radioactive beams. (arXiv:2107.07008v1 [physics.ins-det])

The design, construction, and characterization of the Multi-Sampling Ionization Chamber, MuSIC@Indiana, are described. This detector provides efficient and accurate measurement of the fusion cross-section at near-barrier energies. The response of the detector to low-intensity beams of $^{17,18}$O, $^{19}$F, $^{23}$Na, $^{24,26}$Mg, $^{27}$Al, and $^{28}$Si at E$_{lab}$ = 50-60 MeV was examined. MuSIC@Indiana was commissioned by measuring the $^{18}$O+$^{12}$C fusion excitation function for 11 $15.07.2021 04:26 ATTACC the Quadratic Bottleneck of Attention Layers. (arXiv:2107.06419v1 [cs.LG]) Attention mechanisms form the backbone of state-of-the-art machine learning models for a variety of tasks. Deploying them on deep neural network (DNN) accelerators, however, is prohibitively challenging especially under long sequences. Operators in attention layers exhibit limited reuse and quadratic growth in memory footprint, leading to severe memory-boundedness. This paper introduces a new attention-tailored dataflow, termed FLAT, which leverages operator fusion, loop-nest optimizations, and interleaved execution. It increases the effective memory bandwidth by efficiently utilizing the high-bandwidth, low-capacity on-chip buffer and thus achieves better run time and compute resource utilization. We term FLAT-compatible accelerators ATTACC. In our evaluation, ATTACC achieves 1.94x and 1.76x speedup and 49% and 42% of energy reduction comparing to state-of-the-art edge and cloud accelerators. 14.07.2021 19:44 Scientists develop a new tool for measuring radio waves in fusion plasmas Scientists seeking to bring to Earth the fusion energy that drives the sun and stars use radio frequency (RF) waves—the same waves that bring radio and television into homes—to heat and drive current in the plasma that fuels fusion reactions. Scientists now have developed a path-setting way to measure the waves that could be used to validate predictions of their impact, setting the stage for enhanced future experiments that could result in bringing energy from fusion to Earth. 07:11 Attention-Guided Progressive Neural Texture Fusion for High Dynamic Range Image Restoration. (arXiv:2107.06211v1 [eess.IV]) High Dynamic Range (HDR) imaging via multi-exposure fusion is an important task for most modern imaging platforms. In spite of recent developments in both hardware and algorithm innovations, challenges remain over content association ambiguities caused by saturation, motion, and various artifacts introduced during multi-exposure fusion such as ghosting, noise, and blur. In this work, we propose an Attention-guided Progressive Neural Texture Fusion (APNT-Fusion) HDR restoration model which aims to address these issues within one framework. An efficient two-stream structure is proposed which separately focuses on texture feature transfer over saturated regions and multi-exposure tonal and texture feature fusion. A neural feature transfer mechanism is proposed which establishes spatial correspondence between different exposures based on multi-scale VGG features in the masked saturated HDR domain for discriminative contextual clues over the ambiguous image areas. A progressive texture 06:48 Electrostatic gyrokinetic simulations in Wendelstein 7-X geometry: benchmark between the codes stella and GENE. (arXiv:2107.06060v1 [physics.plasm-ph]) The first experimental campaigns have proven that, due to the optimization of the magnetic configuration with respect to neoclassical transport, the contribution of turbulence is essential to understand and predict the total particle and energy transport in Wendelstein 7-X (W7-X). This has spurred much work on gyrokinetic modelling for the interpretation of the available experimental results and for the preparation of the next campaigns. At the same time, new stellarator gyrokinetic codes have just been or are being developed. It is therefore desirable to have a sufficiently complete, documented and verified set of gyrokinetic simulations in W7-X geometry against which new codes or upgrades of existing codes can be tested and benchmarked. This paper attemps to provide such a set of simulations in the form of a comprehensive benchmark between the recently developed code stella and the well-established code GENE. The benchmark consists of electrostatic gyrokinetic simulations in W7-X 13.07.2021 09:47 Discovery of 10 faces of plasma leads to new insights in fusion and plasma science Scientists have discovered a novel way to classify magnetized plasmas that could possibly lead to advances in harvesting on Earth the fusion energy that powers the sun and stars. The discovery by theorists at the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL) found that a magnetized plasma has 10 unique phases and the transitions between them might hold rich implications for practical development. 12.07.2021 05:41 Multi-level Stress Assessment from ECG in a Virtual Reality Environment using Multimodal Fusion. (arXiv:2107.04566v1 [cs.LG]) ECG is an attractive option to assess stress in serious Virtual Reality (VR) applications due to its non-invasive nature. However, the existing Machine Learning (ML) models perform poorly. Moreover, existing studies only perform a binary stress assessment, while to develop a more engaging biofeedback-based application, multi-level assessment is necessary. Existing studies annotate and classify a single experience (e.g. watching a VR video) to a single stress level, which again prevents design of dynamic experiences where real-time in-game stress assessment can be utilized. In this paper, we report our findings on a new study on VR stress assessment, where three stress levels are assessed. ECG data was collected from 9 users experiencing a VR roller coaster. The VR experience was then manually labeled in 10-seconds segments to three stress levels by three raters. We then propose a novel multimodal deep fusion model utilizing spectrogram and 1D ECG that can provide a stress prediction 05:41 Scaling Gaussian Processes with Derivative Information Using Variational Inference. (arXiv:2107.04061v1 [cs.LG]) Gaussian processes with derivative information are useful in many settings where derivative information is available, including numerous Bayesian optimization and regression tasks that arise in the natural sciences. Incorporating derivative observations, however, comes with a dominating$O(N^3D^3)$computational cost when training on$N$points in$D$input dimensions. This is intractable for even moderately sized problems. While recent work has addressed this intractability in the low-$D$setting, the high-$N$, high-$D$setting is still unexplored and of great value, particularly as machine learning problems increasingly become high dimensional. In this paper, we introduce methods to achieve fully scalable Gaussian process regression with derivatives using variational inference. Analogous to the use of inducing values to sparsify the labels of a training set, we introduce the concept of inducing directional derivatives to sparsify the partial derivative information of a training set. 05:29 Scaling Gaussian Processes with Derivative Information Using Variational Inference. (arXiv:2107.04061v1 [cs.LG]) Gaussian processes with derivative information are useful in many settings where derivative information is available, including numerous Bayesian optimization and regression tasks that arise in the natural sciences. Incorporating derivative observations, however, comes with a dominating$O(N^3D^3)$computational cost when training on$N$points in$D$input dimensions. This is intractable for even moderately sized problems. While recent work has addressed this intractability in the low-$D$setting, the high-$N$, high-$D$setting is still unexplored and of great value, particularly as machine learning problems increasingly become high dimensional. In this paper, we introduce methods to achieve fully scalable Gaussian process regression with derivatives using variational inference. Analogous to the use of inducing values to sparsify the labels of a training set, we introduce the concept of inducing directional derivatives to sparsify the partial derivative information of a training set. 05:29 Free-moving Quantitative Gamma-ray Imaging. (arXiv:2107.04080v1 [physics.ins-det]) The ability to map and estimate the activity of radiological source distributions in unknown three-dimensional environments has applications in the prevention and response to radiological accidents or threats as well as the enforcement and verification of international nuclear non-proliferation agreements. Such a capability requires well-characterized detector response functions, accurate time-dependent detector position and orientation data, an algorithmic understanding of the surrounding 3D environment, and appropriate image reconstruction and uncertainty quantification methods. We have previously demonstrated 3D mapping of gamma-ray emitters with free-moving detector systems on a relative intensity scale using a technique called Scene Data Fusion (SDF). Here we characterize the detector response of a multi-element gamma-ray imaging system using experimentally benchmarked Monte Carlo simulations and perform 3D mapping on an absolute intensity scale. We present experimental 10.07.2021 00:31 Fission vs. fusion: What's the difference? A short overview of two different nuclear processes: fission and fusion. 09.07.2021 06:31 ComFormer: Code Comment Generation via Transformer and Fusion Method-based Hybrid Code Representation. (arXiv:2107.03644v1 [cs.SE]) Developers often write low-quality code comments due to the lack of programming experience, which can reduce the efficiency of developers program comprehension. Therefore, developers hope that code comment generation tools can be developed to illustrate the functionality and purpose of the code. Recently, researchers mainly model this problem as the neural machine translation problem and tend to use deep learning-based methods. In this study, we propose a novel method ComFormer based on Transformer and fusion method-based hybrid code presentation. Moreover, to alleviate OOV (out-of-vocabulary) problem and speed up model training, we further utilize the Byte-BPE algorithm to split identifiers and Sim_SBT method to perform AST Traversal. We compare ComFormer with seven state-of-the-art baselines from code comment generation and neural machine translation domains. Comparison results show the competitiveness of ComFormer in terms of three performance measures. Moreover, we perform a human 07.07.2021 04:48 Hyperspectral Pansharpening Based on Improved Deep Image Prior and Residual Reconstruction. (arXiv:2107.02630v1 [cs.CV]) Hyperspectral pansharpening aims to synthesize a low-resolution hyperspectral image (LR-HSI) with a registered panchromatic image (PAN) to generate an enhanced HSI with high spectral and spatial resolution. Recently proposed HS pansharpening methods have obtained remarkable results using deep convolutional networks (ConvNets), which typically consist of three steps: (1) up-sampling the LR-HSI, (2) predicting the residual image via a ConvNet, and (3) obtaining the final fused HSI by adding the outputs from first and second steps. Recent methods have leveraged Deep Image Prior (DIP) to up-sample the LR-HSI due to its excellent ability to preserve both spatial and spectral information, without learning from large data sets. However, we observed that the quality of up-sampled HSIs can be further improved by introducing an additional spatial-domain constraint to the conventional spectral-domain energy function. We define our spatial-domain constraint as the$L_1$distance between the 04:48 Energy-Efficient Accelerator Design for Deformable Convolution Networks. (arXiv:2107.02547v1 [cs.AR]) Deformable convolution networks (DCNs) proposed to address the image recognition with geometric or photometric variations typically involve deformable convolution that convolves on arbitrary locations of input features. The locations change with different inputs and induce considerable dynamic and irregular memory accesses which cannot be handled by classic neural network accelerators (NNAs). Moreover, bilinear interpolation (BLI) operation that is required to obtain deformed features in DCNs also cannot be deployed on existing NNAs directly. Although a general purposed processor (GPP) seated along with classic NNAs can process the deformable convolution, the processing on GPP can be extremely slow due to the lack of parallel computing capability. To address the problem, we develop a DCN accelerator on existing NNAs to support both the standard convolution and deformable convolution. Specifically, for the dynamic and irregular accesses in DCNs, we have both the input and output 04:37 Single-stage gradient-based stellarator coil design: stochastic optimization. (arXiv:2106.12137v1 [math.OC] CROSS LISTED) We extend the single-stage stellarator coil design approach for quasi-symmetry on axis from [Giuliani et al, 2020] to additionally take into account coil manufacturing errors. By modeling coil errors independently from the coil discretization, we have the flexibility to consider realistic forms of coil errors. The corresponding stochastic optimization problems are formulated using a risk-neutral approach and risk-averse approaches. We present an efficient, gradient-based descent algorithm which relies on analytical derivatives to solve these problems. In a comprehensive numerical study, we compare the coil designs resulting from deterministic and risk-neutral stochastic optimization and find that the risk-neutral formulation results in more robust configurations and reduces the number of local minima of the optimization problem. We also compare deterministic and risk-neutral approaches in terms of quasi-symmetry on and away from the magnetic axis, and in terms of the confinement of 04:37 The rapid destruction of toroidal magnetic surfaces. (arXiv:2107.02717v1 [physics.plasm-ph]) An ideal magnetic evolution can cause the development of an exponentially large variation between the distance of closest approach and greatest separation between neighboring pairs of magnetic field lines. When this occurs, a fast magnetic reconnection naturally arises on the evolution time scale of magnetic field times a factor that depends only logarithmically on the strength of the non-ideal effects. An obvious example arises when the magnetic evolution is driven by footpoint motion, as in the solar corona. A similar effect can be responsible for the sudden loss of magnetic surfaces during a tokamak disruption. In almost all magnetic surfaces in toroidal geometry, a magnetic field line never closes on itself as it is followed in the toroidal angle$\varphi$, and the line comes arbitrarily close to every point in the surface. When an arbitrary pair of magnetic field lines are separated by an infinitesimal distance$\delta_0$in a surface at$\varphi=0$, then their separation can be 06.07.2021 07:28 Stellarator coil design using cubic splines for improved access on the outboard side. (arXiv:2107.02123v1 [physics.plasm-ph]) In recent years many efforts have been undertaken to simplify coil designs for stellarators due to the difficulties in fabricating non-planar coils. The FOCUS code removes the need for a winding surface and represents the coils as arbitrary curves in 3D. In the following work, the implementation of a spline representation for the coils in FOCUS is described, along with the implementation of a new engineering constraint to design coils with a straighter outer section. The new capabilities of the code are shown as an example on HSX, NCSX, and a prototype quasi-axisymmetric reactor-sized stellarator. The flexibility granted by splines along with the new constraint will allow for stellarator coil designs with improved accessibility and simplified maintenance 07:28 Influence of rotation on axisymmetric plasma equilibria: double-null DTT scenario. (arXiv:2107.01890v1 [physics.plasm-ph]) We study the dependence of some relevant tokamak equilibrium quantities on the toroidal plasma rotation. The Grad-Shafranov equation generalised to the rotating case is analytically solved employing two different representations for the homogenous solution. Using an expression in terms of polynomials, we describe the separatrix shape by a few geometrical parameters, reproducing different plasma scenarios such as double-null and inverse triangularity. In this setting, the introduction of toroidal rotation corresponds to variations on relevant plasma quantities, most notably an enhancement of the poloidal beta. Using a more general expression in terms of Bessel functions, we reconstruct the full plasma boundary of the double-null configuration proposed for the upcoming DTT experiment, demonstrating how said configuration is compatible with different values of the plasma velocity. 02.07.2021 04:25 Attention Bottlenecks for Multimodal Fusion. (arXiv:2107.00135v1 [cs.CV]) Humans perceive the world by concurrently processing and fusing high-dimensional inputs from multiple modalities such as vision and audio. Machine perception models, in stark contrast, are typically modality-specific and optimised for unimodal benchmarks, and hence late-stage fusion of final representations or predictions from each modality (late-fusion') is still a dominant paradigm for multimodal video classification. Instead, we introduce a novel transformer based architecture that uses fusion bottlenecks' for modality fusion at multiple layers. Compared to traditional pairwise self-attention, our model forces information between different modalities to pass through a small number of bottleneck latents, requiring the model to collate and condense the most relevant information in each modality and only share what is necessary. We find that such a strategy improves fusion performance, at the same time reducing computational cost. We conduct thorough ablation studies, and achieve 30.06.2021 09:48 Nonlinear dynamics and phase space transport by chorus emission. (arXiv:2106.15026v1 [physics.plasm-ph]) Chorus emission in planetary magnetospheres is taken as working paradigm to motivate a short tutorial trip through theoretical plasma physics methods and their applications. Starting from basic linear theory, readers are first made comfortable with whistler wave packets and their propagation in slowly varying weakly nonuniform media, such as the Earth's magnetosphere, where they can be amplified by a population of supra-thermal electrons. The nonlinear dynamic description of energetic electrons in the phase space in the presence of self-consistently evolving whistler fluctuation spectrum is progressively introduced by addressing renormalization of the electron response and spectrum evolution equations. Analytical and numerical results on chorus frequency chirping are obtained and compared with existing observations and particle in cell simulations. Finally, the general theoretical framework constructed during this short trip through chorus physics is used to draw analogies with 09:48 Equations and improved coefficients for paralleltransport in multicomponent collisional plasmas: method and application for tokamak modelling. (arXiv:2106.14933v1 [physics.plasm-ph]) New analytical expressions for parallel transport coefficients in multicomponent collisional plasmas are presented in this paper. They are improved versions of the expressions written in [V. M. Zhdanov. Transport Processes in Multicomponent Plasma, vol. 44. 10 2002.], based on Grad's 21N-moment method. Both explicit and approximate approaches for transport coefficients calculation are considered. Accurate application of this closure for the Braginskii transport equations is discussed. Viscosity dependence on the heat flux is taken into account. Improved expressions are implemented into the SOLPS-ITER code and tested for deuterium and neon ITER cases. Some typos found in [V. M. Zhdanov. Transport Processes in Multicomponent Plasma, vol. 44. 10 2002.] are corrected. 09:48 Stellarator optimization for good magnetic surfaces at the same time as quasisymmetry. (arXiv:2106.14930v1 [physics.plasm-ph]) A method is demonstrated to optimize a stellarator's geometry to eliminate magnetic islands and achieve other desired physics properties at the same time. For many physics quantities that have been used in stellarator optimization, including quasisymmetry, neoclassical transport, and magnetohydrodynamic stability, it is convenient to use a magnetic equilibrium representation that assures the existence of magnetic surfaces. However, this representation hides the possible presence of magnetic islands, which are typically undesirable. To include both surface-based objectives and island widths in a single optimization, two fixed-boundary equilibrium calculations are run at each iteration of the optimization: one that enforces the existence of magnetic surfaces (VMEC [S. P. Hirshman and J. C. Whitson, Phys. Fluids 26, 3553 (1983)]), and one that does not (SPEC [S. R. Hudson, et al, Phys. Plasmas 19, 112502 (2012)]). By penalizing the island residues in the objective function, the two 29.06.2021 05:38 Modeling of carbon pellets disruption mitigation in an NSTX-U plasma. (arXiv:2106.14788v1 [physics.plasm-ph]) Single carbon pellet disruption mitigation simulations using M3D-C1 were conducted in an NSTX-U-like plasma to support the electromagnetic pellet injection concept (EPI). A carbon ablation model has been implemented in M3D-C1 and tested with available data. 2D simulations were conducted in order to estimate the amount of carbon needed to quench the plasma, finding that the content in a$1\,$mm radius vitreous carbon pellet (~ 3.2x10E20 atoms) would be enough if it is entirely ablated. 3D simulations were performed, scanning over pellet velocity and parallel thermal conductivity, as well as different injection directions and pellet concepts (solid pellets and shell pellets). The sensitivity of the thermal quench and other related quantities to these parameters has been evaluated. A 1 mm radius solid pellet only partially ablates at velocities of 300 m/s or higher, thus being unable to fully quench the plasma. To further enhance the ablation, approximations to an array of pellets and the 05:38 OCCAM: Optimal Data Reuse for Convolutional Neural Networks. (arXiv:2106.14138v1 [cs.AR]) Convolutional neural networks (CNNs) are emerging as powerful tools for image processing in important commercial applications. We focus on the important problem of improving the latency of image recognition. CNNs' large data at each layer's input, filters, and output poses a memory bandwidth problem. While previous work captures only some of the enormous data reuse, full reuse implies that the initial input image and filters are read once from off chip and the final output is written once off chip without spilling the intermediate layers' data to off-chip. We propose Occam to capture full reuse via four contributions. (1) We identify the necessary condition for full reuse. (2) We identify the dependence closure as the sufficient condition to capture full reuse using the least on-chip memory. (3) Because the dependence closure is often too large to fit in on-chip memory, we propose a dynamic programming algorithm that optimally partitions a given CNN to guarantee the least off-chip 28.06.2021 11:40 AI Enables Fast, Accurate Predictions for Better Control of Fusion Experiments The software that drives self-driving cars and digital assistants is machine learning, a method employed in artificial intelligence (AI). It currently allows researchers to tackle the main challenges... 08:16 Feasibility Study and Perspectives of proton Dielectric Laser Accelerators (p-DLA): from nanosource to accelerator scheme. (arXiv:2106.13701v1 [physics.acc-ph]) In this paper we discuss the possibility to generate and accelerate proton nanobeams in fully dielectric laser-driven accelerators (p-DLAs). High gradient on-chip optical-power dielectric laser accelerators (DLAs) could represent one of the most promising way towards future miniaturized particle accelerator. A primary challenge for DLAs are small beam apertures having a size of the order of the driving laser wavelength where low charge high-repetition (or also CW) ultralow emittance nanobeams have to be transported. For electrons beams generation and acceleration, intense research activities are ongoing, and several demonstrations have been already obtained by using electrons nanotip (or flat photocathode) sources feeding dielectric microstructures. In this article we aim at the possibility to integrate a nanosource for the generation of a light ion or proton nano-beams suitable for the subsequent acceleration into sub-relativistic (low-beta) p-DLA stages. Such integration includes the 08:16 The Notre-Dame Cube: An active-target time-projection chamber for radioactive beam experiments and detector development. (arXiv:2106.13236v1 [physics.ins-det]) Active-target detectors have the potential to address the difficulties associated with the low intensities of radioactive beams. We have developed an active-target detector, the Notre Dame Cube (ND-Cube), to perform experiments with radioactive beams produced at$\mathit{TwinSol}$and to aid in the development of active-target techniques. Various aspects of the ND-Cube and its design were characterized. The ND-Cube was commissioned with a$^{7}$Li beam for measuring$^{40}$Ar +$^{7}$Li fusion reaction cross sections and investigating$^{7}$Li($\alpha$,$\alpha$)$^{7}$Li scattering events. The ND-Cube will be used to study a range of reactions using light radioactive ions produced at low energy. 25.06.2021 09:30 Identifying Hidden Visits from Sparse Call Detail Record Data. (arXiv:2106.12885v1 [stat.AP]) Despite a large body of literature on trip inference using call detail record (CDR) data, a fundamental understanding of their limitations is lacking. In particular, because of the sparse nature of CDR data, users may travel to a location without being revealed in the data, which we refer to as a "hidden visit". The existence of hidden visits hinders our ability to extract reliable information about human mobility and travel behavior from CDR data. In this study, we propose a data fusion approach to obtain labeled data for statistical inference of hidden visits. In the absence of complementary data, this can be accomplished by extracting labeled observations from more granular cellular data access records, and extracting features from voice call and text messaging records. The proposed approach is demonstrated using a real-world CDR dataset of 3 million users from a large Chinese city. Logistic regression, support vector machine, random forest, and gradient boosting are used to infer 09:19 Identifying Hidden Visits from Sparse Call Detail Record Data. (arXiv:2106.12885v1 [stat.AP]) Despite a large body of literature on trip inference using call detail record (CDR) data, a fundamental understanding of their limitations is lacking. In particular, because of the sparse nature of CDR data, users may travel to a location without being revealed in the data, which we refer to as a "hidden visit". The existence of hidden visits hinders our ability to extract reliable information about human mobility and travel behavior from CDR data. In this study, we propose a data fusion approach to obtain labeled data for statistical inference of hidden visits. In the absence of complementary data, this can be accomplished by extracting labeled observations from more granular cellular data access records, and extracting features from voice call and text messaging records. The proposed approach is demonstrated using a real-world CDR dataset of 3 million users from a large Chinese city. Logistic regression, support vector machine, random forest, and gradient boosting are used to infer 09:19 Enhanced D-D Fusion Rates when the Coulomb Barrier Is Lowered by Electrons. (arXiv:2106.12988v1 [physics.plasm-ph]) A profusion of unbound, low-energy electrons creates a local electric field that reduces Coulomb potential and increases quantum tunneling probability for pairs of nuclei. Neutral beam-target experiments on deuterium-deuterium fusion reactions, observed with neutron detectors, show percentage increases in fusion products are consistent with electron-screening predictions from Schrodinger wave mechanics. Experiments performed confirm that observed fusion rate enhancement with a negatively biased target is primarily due to changes to the fusion cross section, rather than simply acceleration due to electrostatic forces. 09:19 Thermal Quench in ITER Locked Mode Disruptions. (arXiv:2106.12943v1 [physics.plasm-ph]) Simulations and theory are presented of an ITER locked mode thermal quench (TQ). In present experiments, locked mode disruptions have a long precursor phase, followed by a rapid termination and thermal quench, which can be identified with a resistive wall tearing mode (RWTM). In ITER, the RWTM will be slowed by the highly conductive vacuum vessel. The rapid termination might be absent, and the plasma could remain in the precursor phase. If the edge temperature is in the collisional regime, the TQ would proceed on a long timescale, limited by the RWTM to almost$100ms.This is an important self mitigating effect. 24.06.2021 23:26 Artificial intelligence speeds forecasts to control fusion experiments Machine learning, a technique used in the artificial intelligence (AI) software behind self-driving cars and digital assistants, now enables scientists to address key challenges to harvesting on Earth the fusion energy that powers the sun and stars. The technique recently empowered physicist Dan Boyer of the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL) to develop fast and accurate predictions for advancing control of experiments in the National Spherical Torus Experiment-Upgrade (NSTX-U)—the flagship fusion facility at PPPL that is currently under repair. 05:13 Signals to Spikes for Neuromorphic Regulated Reservoir Computing and EMG Hand Gesture Recognition. (arXiv:2106.11169v2 [eess.SP] UPDATED) Surface electromyogram (sEMG) signals result from muscle movement and hence they are an ideal candidate for benchmarking event-driven sensing and computing. We propose a simple yet novel approach for optimizing the spike encoding algorithm's hyper-parameters inspired by the readout layer concept in reservoir computing. Using a simple machine learning algorithm after spike encoding, we report performance higher than the state-of-the-art spiking neural networks on two open-source datasets for hand gesture recognition. The spike encoded data is processed through a spiking reservoir with a biologically inspired topology and neuron model. When trained with the unsupervised activity regulation CRITICAL algorithm to operate at the edge of chaos, the reservoir yields better performance than state-of-the-art convolutional neural networks. The reservoir performance with regulated activity was found to be 89.72% for the Roshambo EMG dataset and 70.6% for the EMG subset of sensor fusion dataset. 05:02 Single-stage gradient-based stellarator coil design: stochastic optimization. (arXiv:2106.12137v1 [math.OC]) We extend the single-stage stellarator coil design approach for quasi-symmetry on axis from [Giuliani et al, 2020] to additionally take into account coil manufacturing errors. By modeling coil errors independently from the coil discretization, we have the flexibility to consider realistic forms of coil errors. The corresponding stochastic optimization problems are formulated using a risk-neutral approach and risk-averse approaches. We present an efficient, gradient-based descent algorithm which relies on analytical derivatives to solve these problems. In a comprehensive numerical study, we compare the coil designs resulting from deterministic and risk-neutral stochastic optimization and find that the risk-neutral formulation results in more robust configurations and reduces the number of local minima of the optimization problem. We also compare deterministic and risk-neutral approaches in terms of quasi-symmetry on and away from the magnetic axis, and in terms of the confinement of 23.06.2021 07:52 BEyond observation: an approach for ObjectNav. (arXiv:2106.11379v1 [cs.CV]) With the rise of automation, unmanned vehicles became a hot topic both as commercial products and as a scientific research topic. It composes a multi-disciplinary field of robotics that encompasses embedded systems, control theory, path planning, Simultaneous Localization and Mapping (SLAM), scene reconstruction, and pattern recognition. In this work, we present our exploratory research of how sensor data fusion and state-of-the-art machine learning algorithms can perform the Embodied Artificial Intelligence (E-AI) task called Visual Semantic Navigation. This task, a.k.a Object-Goal Navigation (ObjectNav) consists of autonomous navigation using egocentric visual observations to reach an object belonging to the target semantic class without prior knowledge of the environment. Our method reached fourth place on the Habitat Challenge 2021 ObjectNav on the Minival phase and the Test-Standard Phase. 07:41 On the genesis and nature of Palm Tree Modes in the JET tokamak. (arXiv:2106.11602v1 [physics.plasm-ph]) Long-lived, highly localized structures called palm tree modes (PTM) are observed in the edge plasma of the JET tokamak. Although PTMs are well documented, little is known about the mechanisms which produce these structures. In the case of the PTM, an ELM-postcursor, its genesis is usually explained by ergodisation of the magnetic field due to edge localized modes and the appearance of a seed magnetic island which evolves into a PTM later. In this study we try to invoke a creation mechanism based on the concepts and observations in edge plasma turbulence. An interesting aspect of plasma turbulence is the occurrence of coherent, long-lived structures in the scrape-off-layer (SOL). These localized and magnetic-field-aligned regions with higher or lower plasma densities are called blobs and holes. Measurements show that these filaments carry parallel currents. We thus here interpret ELM-filaments as massive blobs and the interspace between these filaments as holes. We demonstrate that a 22.06.2021 10:42 Two-Faced Humans on Twitter and Facebook: Harvesting Social Multimedia for Human Personality Profiling. (arXiv:2106.10673v1 [cs.SI]) Human personality traits are the key drivers behind our decision-making, influencing our life path on a daily basis. Inference of personality traits, such as Myers-Briggs Personality Type, as well as an understanding of dependencies between personality traits and users' behavior on various social media platforms is of crucial importance to modern research and industry applications. The emergence of diverse and cross-purpose social media avenues makes it possible to perform user personality profiling automatically and efficiently based on data represented across multiple data modalities. However, the research efforts on personality profiling from multi-source multi-modal social media data are relatively sparse, and the level of impact of different social network data on machine learning performance has yet to be comprehensively evaluated. Furthermore, there is not such dataset in the research community to benchmark. This study is one of the first attempts towards bridging such an 10:31 A simulation study of a windowless gas stripping room in an E//B neutral particle analyzer. (arXiv:2106.11139v1 [physics.ins-det]) Neutral Particle Analyzer (NPA) is one of the crucial diagnostic devices on Tokamak facilities. Stripping unit is one of the main parts of the NPA. A windowless gas stripping room with two differential pipes is adopted in a parallel direction of electric and magnetic fields (E//B) NPA. The pressure distributions in the stripping chamber are simulated by Ansys Fluent together with MolFlow+. Based on the pressure distributions extracted from the simulation, the stripping efficiency of the E//B NPA is studied with GEANT4. The hadron reaction physics is modified to track the charge state of each particle in a cross section base method in GEANT4. The transmission rates (R$) and the stripping efficiencies$f_{+1}$are examined for the particle energy ranging from 20 to 200 keV at the input pressure ($P_0$) ranging from 20 to 400 Pa. According to the combined global efficiency,$R \times f_{+1}$,$P_0$= 240 Pa is obtained as the optimum pressure for the maximum global efficiency in the 21.06.2021 06:18 Optimising simulations for diphoton production at hadron colliders using amplitude neural networks. (arXiv:2106.09474v1 [hep-ph] CROSS LISTED) Machine learning technology has the potential to dramatically optimise event generation and simulations. We continue to investigate the use of neural networks to approximate matrix elements for high-multiplicity scattering processes. We focus on the case of loop-induced diphoton production through gluon fusion and develop a realistic simulation method that can be applied to hadron collider observables. Neural networks are trained using the one-loop amplitudes implemented in the NJet C++ library and interfaced to the Sherpa Monte Carlo event generator where we perform a detailed study for$2\to3$and$2\to4$scattering problems. We also consider how the trained networks perform when varying the kinematic cuts effecting the phase space and the reliability of the neural network simulations. 17.06.2021 17:54 Fusion of artificial intelligence and nanopore technology enables rapid point-of-care COVID test Scientists have demonstrated that single virus particles passing through a nanopore could be accurately identified using machine learning. The test platform they created was so sensitive that the coronaviruses responsible for the common cold, SARS, MERS, and COVID could be distinguished from each other. 16:53 Intelligent Tire-Based Slip Ratio Estimation Using Different Machine Learning Algorithms. (arXiv:2106.08961v1 [cs.LG]) Estimation of the longitudinal slip ratio of tires is important in boosting the control performance of the vehicle under driving and braking conditions. In this paper, the slip ratio is estimated using four machine learning algorithms (Neural Network, Gradient Boosting Machine, Random Forest and Support Vector Machine) based on the acceleration signals from the tri-axial MEMS accelerometers utilized in the intelligent tire system. The experimental data are collected through the MTS experimental platform. The corresponding acceleration signals within the tire contact patch are extracted after filtering to be used for the training the aforesaid machine learning algorithms. A comparison is provided between the implemented ML algorithms using a 10-fold CV. NRMS errors in the CV results indicate that NN has the highest accuracy in comparison with other techniques. The NRSM errors of NN, GBM, RF, and SVM are 2.59\%, 3.30\%, 4.21\%, and 5.34\%, respectively. Among these techniques, GBM has a 03:52 Plans unveiled for private U.K. fusion reactor powered by ‘smoke rings’ and pneumatic pistons Canadian startup General Fusion says pilot plant could turn on in 2025 16.06.2021 05:17 Design and Testing of Dimes Carbon Ablation Rods in the DIII-D Tokamak. (arXiv:2106.08306v1 [physics.plasm-ph]) We present the design of ATJ graphite rods developed for ablation experiments under high heat flux (up to 50 MW/m2) in the lower divertor of the DIII-D tokamak [1], a magnetic plasma confinement device. This work is motivated by the need to test ablation models relevant to carbon-based thermal shields used in high-speed spacecraft atmospheric entries, where the heat fluxes encountered can be comparable to those achieved in the DIII-D divertor plasma. Several different designs for the flow-facing side of the rod are analyzed, including "sharp nose," "blunt," and "concave". The last shape is studied for its potential to lower heat fluxes at the rod surface by increased radiation from trapped neutrals and reduced parallel plasma pressure. We also analyze the possibility of applying a thin (approximately 30 microns) layer of silicon carbide (SiC) to the exposed part of several carbon ablation rods to benchmark its erosion calculations and lifetime predictions. Such calculations are of 05:17 Ponderomotive Screening of Nuclear Fusion Reactions Based on Localized Surface Plasmon Resonance. (arXiv:2106.08127v1 [physics.gen-ph]) A scheme is presented to catalyze nuclear fusion by the excitation of the localized surface plasmon resonance (LSPR) of an Au nanobipyramid (NBP) by ultrafast (femtosecond) laser pulses. The effect utilizes the exceptionally high electric field enhancement provided by the Au NBP LSPR, localized to nm^3 volumes and fs time scales, to produce a ponderomotive force that in turn provides for screening energies near the Au NBP tip. Discrete-dipole approximation (DDA) simulations are presented to support the proposed scheme. Calculations made with conservative parameters suggest that the effect should be observable in a laboratory setting using commercially available ultrafast lasers. 15.06.2021 20:24 World's most powerful magnet ready to ship After a decade of design and fabrication, General Atomics is ready to ship the first module of the Central Solenoid, the world's most powerful magnet. It will become a central component of ITER, a machine that replicates the fusion power of the sun. ITER is being built in southern France by 35 partner countries. 20:08 World's most powerful magnet being shipped to ITER fusion reactor The world’s most powerful magnet, 280,000 times stronger than Earth's own magnetic field, is being shipped to France for installation in the core of the ITER fusion reactor 19:17 World's most powerful magnet begins journey to heart of giant fusion experiment Engineers are preparing to move part of the world's most powerful magnet from the U.S. to France as a part of an international nuclear fusion project. 08:05 Low-threshold induced side-scattering instability in the edge transport barrier at O-mode ECRH experiments in magnetic fusion devices. (arXiv:2106.07489v1 [physics.plasm-ph]) The possibility of electron plasma waves trapping in the steepest part of the tokamak edge transport barrier leading to the excitation of low-threshold side-scattering absolute parametric decay instability in ordinary mode electron cyclotron resonance heating (ECRH) experiments is demonstrated. The instability is shown to be important for ECRH experiments in ITER. 08:05 The surprising attractiveness of tearing mode locking in tokamaks. (arXiv:2106.06581v1 [physics.plasm-ph]) Tearing modes in tokamaks typically rotate while small and then lock at a fixed location when larger. Research on present-day devices has focused almost exclusively on stabilisation of rotating modes, as it has been considered imperative to avoid locked modes. However, in larger devices, such as those contemplated for tokamak reactors, the locking occurs at a smaller island size, and the island can be safely stabilised after locking. The stabilisation of small locked modes can be performed at lower wave power and broader deposition compared to rotating islands. On large devices, it thus becomes surprisingly advantageous to allow the mode to grow and lock naturally before stabilising it. Calculations indicate that the ITER international megaproject would be best stabilised through this approach. 14.06.2021 10:41 Using HPC infrastructures for deep learning applications in fusion research. (arXiv:2106.06101v1 [physics.comp-ph]) In the fusion community, the use of high performance computing (HPC) has been mostly dominated by heavy-duty plasma simulations, such as those based on particle-in-cell and gyrokinetic codes. However, there has been a growing interest in applying machine learning for knowledge discovery on top of large amounts of experimental data collected from fusion devices. In particular, deep learning models are especially hungry for accelerated hardware, such as graphics processing units (GPUs), and it is becoming more common to find those models competing for the same resources that are used by simulation codes, which can be either CPU- or GPU-bound. In this paper, we give examples of deep learning models -- such as convolutional neural networks, recurrent neural networks, and variational autoencoders -- that can be used for a variety of tasks, including image processing, disruption prediction, and anomaly detection on diagnostics data. In this context, we discuss how deep learning can go from 10:41 ViT-Inception-GAN for Image Colourising. (arXiv:2106.06321v1 [cs.CV]) Studies involving colourising images has been garnering researchers' keen attention over time, assisted by significant advances in various Machine Learning techniques and compute power availability. Traditionally, colourising images have been an intricate task that gave a substantial degree of freedom during the assignment of chromatic information. In our proposed method, we attempt to colourise images using Vision Transformer - Inception - Generative Adversarial Network (ViT-I-GAN), which has an Inception-v3 fusion embedding in the generator. For a stable and robust network, we have used Vision Transformer (ViT) as the discriminator. We trained the model on the Unsplash and the COCO dataset for demonstrating the improvement made by the Inception-v3 embedding. We have compared the results between ViT-GANs with and without Inception-v3 embedding. 11.06.2021 04:52 A model for the fast evaluation of prompt losses of energetic ions in stellarators. (arXiv:2106.05697v1 [physics.plasm-ph]) A good understanding of the confinement of energetic ions in non-axisymmetric magnetic fields is key for the design of reactors based on the stellarator concept. In this work, we develop a model that, based on the radially-local bounce-averaged drift-kinetic equation, classifies orbits and succeeds in predicting configuration-dependent aspects of the prompt losses of energetic ions in stellarators. Such a model could in turn be employed in the optimization stage of the design of new devices. 04:52 A neoclassically optimized compact stellarator with four planar coils. (arXiv:2106.05576v1 [physics.plasm-ph]) A neoclassically optimized compact stellarator with simple coils has been designed. The magnetic field of the new stellarator is generated by only four planar coils including two interlocking coils of elliptical shape and two circular poloidal field coils. The interlocking coil topology is the same as that of the Columbia Non-neutral Torus (CNT). The new configuration was obtained by minimizing the effective helical ripple directly via the shape of the two interlocking coils. The optimized compact stellarator has very low effective ripple in the plasma core implying excellent neoclassical confinement. This is confirmed by the results of the drift-kinetic code SFINCS showing that the particle diffusion coefficient of the new configuration is one order of magnitude lower than CNT's. 04:52 A concise method for feature selection via normalized frequencies. (arXiv:2106.05814v1 [cs.LG]) Feature selection is an important part of building a machine learning model. By eliminating redundant or misleading features from data, the machine learning model can achieve better performance while reducing the demand on com-puting resources. Metaheuristic algorithms are mostly used to implement feature selection such as swarm intelligence algorithms and evolutionary algorithms. However, they suffer from the disadvantage of relative complexity and slowness. In this paper, a concise method is proposed for universal feature selection. The proposed method uses a fusion of the filter method and the wrapper method, rather than a combination of them. In the method, one-hoting encoding is used to preprocess the dataset, and random forest is utilized as the classifier. The proposed method uses normalized frequencies to assign a value to each feature, which will be used to find the optimal feature subset. Furthermore, we propose a novel approach to exploit the outputs of mutual information, 10.06.2021 15:51 Astronomers probe layer-cake structure of brown dwarf's atmosphere Brown dwarfs are the cosmic equivalent of tweeners. They're too massive to be planets and too small to sustain nuclear fusion in their cores, which powers stars. Many brown dwarfs are nomadic. They do not orbit stars but drift among them as loners. 07:39 Turbulent transport of impurities in 3D devices. (arXiv:2106.05017v1 [physics.plasm-ph]) A large diffusive turbulent contribution to the radial impurity transport in Wendelstein 7-X (W7-X) plasmas has been experimentally inferred during the first campaigns and numerically confirmed by means of gyrokinetic simulations with the code stella. In general, the absence of strong impurity accumulation during the initial W7-X campaigns is attributed to this diffusive term. In the present work the diffusive contribution is also calculated in other stellarator plasmas. In particular, the diffusion (D) and convection (V) coefficients of carbon and iron impurities produced by ion-temperature-gradient (ITG) turbulence are obtained for W7-X, LHD, TJ-II and NCSX. The results show that, although the size of D and V can differ across the four devices, inward convection is found for all of them. For W7-X, TJ-II and NCSX the two coefficients are comparable and the turbulent peaking factor is surprisingly similar. In LHD, appreciably weaker diffusive and convective impurity transport and 09.06.2021 10:16 Modeling of Particle Transport, Neutrals and Radiation in Magnetically-Confined Plasmas with AURORA. (arXiv:2106.04528v1 [physics.plasm-ph]) We present AURORA, an open-source package for particle transport, neutrals and radiation modeling in magnetic confinement fusion plasmas. AURORA's modern multi-language interface enables simulations of 1.5D impurity transport within high-performance computing frameworks, particularly for the inference of particle transport coefficients. A user-friendly Python library allows simple interaction with atomic rates from the Atomic Data and Atomic Structure database [Summers 2006] as well as other sources. This enables a range of radiation predictions, both for power balance and spectroscopic analysis. We discuss here the superstaging approximation for complex ions, as a way to group charge states and reduce computational cost, demonstrating its wide applicability within the AURORA forward model and beyond. AURORA also facilitates neutral particle analysis, both from experimental spectroscopic data and other simulation codes. Leveraging AURORA's capabilities to interface SOLPS-ITER [Wiesen 10:16 Topological impact of a simple self-replication geometry structure with great application potential in vacuum pumping and photovoltaic industry. (arXiv:2106.04451v1 [physics.app-ph]) Topological effects exist from a macroscopic system such as the universe to a microscopic system described by the quantum mechanics. We show here that an interesting geometry structure can be created by self-replication procedure of a square with an inscribed circle, in which the sum of the circles area will remain the same but the sum of circumference will increase. It is demonstrated that these topological features have significant impacts to the vacuum pumping probability and the photon absorption probability of the active surface by the Monte Carlo simulation. The results show great application potential in vacuum pumping of large research facilities such as nuclear fusion reactor, synchrotron, gravitational-wave detector, and in photovoltaic industry. 10:16 Magnetoconvection in a horizontal duct flow at very high Hartmann and Grashof numbers. (arXiv:2106.04231v1 [physics.flu-dyn]) Direct numerical simulations and linear stability analysis are carried out to study mixed convection in a horizontal duct with constant-rate heating applied at the bottom and imposed transverse horizontal magnetic field. A two-dimensional approximation corresponding to the asymptotic limit of very strong magnetic field effect is validated and applied, together with full three-dimensional analysis, to investigate the flow's behaviour in the previously unexplored range of control parameters corresponding to typical conditions of a liquid metal blanket of a nuclear fusion reactor. It is found that the instability to quasi-two-dimensional rolls parallel to the magnetic field discovered at smaller Hartmann and Grashof numbers in earlier studies also occurs in this parameter range. Transport of the rolls by the mean flow leads to magnetoconvective temperature fluctuations of exceptionally high amplitudes. It is also demonstrated that the quasi-two-dimensional structure of flows at very high 08.06.2021 05:43 BayesIMP: Uncertainty Quantification for Causal Data Fusion. (arXiv:2106.03477v1 [stat.ML]) While causal models are becoming one of the mainstays of machine learning, the problem of uncertainty quantification in causal inference remains challenging. In this paper, we study the causal data fusion problem, where datasets pertaining to multiple causal graphs are combined to estimate the average treatment effect of a target variable. As data arises from multiple sources and can vary in quality and quantity, principled uncertainty quantification becomes essential. To that end, we introduce Bayesian Interventional Mean Processes, a framework which combines ideas from probabilistic integration and kernel mean embeddings to represent interventional distributions in the reproducing kernel Hilbert space, while taking into account the uncertainty within each causal graph. To demonstrate the utility of our uncertainty estimation, we apply our method to the Causal Bayesian Optimisation task and show improvements over state-of-the-art methods. 05:43 A new heat source model for selective laser melting simulations based on energy distribution on the powder layer and the surface of substrate. (arXiv:2106.03482v1 [physics.app-ph]) In order to predict the more accurate shape information of the melt pool in Selective Laser Melting (SLM), a new finite element temperature field simulations model is proposed. The simulations use a new heat source model that takes into account the influence of the powder layout, the surface of the substrate and the changes in the thickness of the powder layer after fusion on the energy distribution. In order to construct this new heat source model, firstly an improved optimization method based on the gradient descent and the univariate search technique is proposed to simulate the powder layout, and then the laser beam propagation between the powder and the surface of the substrate is tracked and recorded to obtain the energy distribution. Finally, according to the distribution of laser energy between the powder layer and the surface of the substrate, the heat source model is divided into two parts: one is the surface of substrate heat source model being the Gaussian distribution, the 05:43 The role of plasma-molecule interactions on power and particle balance during detachment on the TCV tokamak. (arXiv:2106.03430v1 [physics.plasm-ph]) This paper shows experimental results from the TCV tokamak that indicate plasma-molecule interactions involving$D_2^+$and possibly$D^-$play an important role as sinks of energy (through hydrogenic radiation as well as dissociation) and particles during divertor detachment if low target temperatures ($ The impact of $D_2^+$ is shown to be underestimated in present (vibrationally unresolved) SOLPS-ITER simulations, which could result from an underestimated $D_2 + D^+ \rightarrow D_2^+ + D$ rate. The converged SOLPS-ITER

07.06.2021
15:00 China’s Fusion Reactor Sets World Record by Running for 101 Seconds

Chinese state media has reported that EAST has taken a big step toward making fusion power a reality by keeping plasma at 120 million degrees Celsius for 101 seconds. The post China’s Fusion Reactor Sets World Record by Running for 101 Seconds appeared first on ExtremeTech.