Search

In the field of locomotion task of quadruped robots, Blind Policy and Perceptive Policy each have their own advantages and limitations. The Blind Policy relies on preset sensor information and algorithms, suitable for known and structured environments, but it lacks adaptability in complex or unknown environments. The Perceptive Policy uses visual sensors to obtain detailed environmental information, allowing it to adapt to complex terrains, but its effectiveness is limited under occluded conditions, especially when perception fails. Unlike the Blind Policy, the Perceptive Policy is not as robust under these conditions. To address these challenges, we propose a MBC:Multi-Brain collaborative system that incorporates the concepts of Multi-Agent Reinforcement Learning and introduces collaboration between the Blind Policy and the Perceptive Policy. By applying this multi-policy collaborative model to a quadruped robot, the robot can maintain stable locomotion even when the perceptual system is impaired or observational data is incomplete. Our simulations and real-world experiments demonstrate that this system significantly improves the robot's passability and robustness against perception failures in complex environments, validating the effectiveness of multi-policy collaboration in enhancing robotic motion performance., Comment: 18 pages, 9 figures, Website and Videos: https://quad-mbc.github.io/

It was reported that the optical field fluctuations in self-defocusing media can be described by sound waves propagating in a two-dimensional photon-fluid which is controlled by the driving laser beam. The photon-fluid can be regarded as the background where the sound waves propagate in the way like a scalar field propagating in curved spacetime, thus providing a platform to study physics in analog gravity. In this analog gravity model, to be more specific, in the analog rotating black hole background, we study the quasinormal modes (QNMs) of massive scalar field perturbations, as a natural extension of the recent work on the massless QNMs in photon-fluid. We analyze the properties of the spectrum of fundamental QNMs which are calculated by Continued Fraction Method and WKB approximation method. We also investigate the quasi-resonance and find that it may exist in this analog gravity model. The existence of quasi-resonance is important to study the QNMs in analog gravity apparatus due to its slow damping rate and longevity which give us better opportunity to probe and study it in laboratory., Comment: 28 pages, 6 figures

Designing a bipedal robot is a complex and challenging task, especially when dealing with a multitude of structural parameters. Traditional design methods often rely on human intuition and experience. However, such approaches are time-consuming, labor-intensive, lack theoretical guidance and hard to obtain optimal design results within vast design spaces, thus failing to full exploit the inherent performance potential of robots. In this context, this paper introduces the SERL (Structure Evolution Reinforcement Learning) algorithm, which combines reinforcement learning for locomotion tasks with evolution algorithms. The aim is to identify the optimal parameter combinations within a given multidimensional design space. Through the SERL algorithm, we successfully designed a bipedal robot named Wow Orin, where the optimal leg length are obtained through optimization based on body structure and motor torque. We have experimentally validated the effectiveness of the SERL algorithm, which is capable of optimizing the best structure within specified design space and task conditions. Additionally, to assess the performance gap between our designed robot and the current state-of-the-art robots, we compared Wow Orin with mainstream bipedal robots Cassie and Unitree H1. A series of experimental results demonstrate the Outstanding energy efficiency and performance of Wow Orin, further validating the feasibility of applying the SERL algorithm to practical design.

Optical remote sensing and Synthetic Aperture Radar(SAR) remote sensing are crucial for earth observation, offering complementary capabilities. While optical sensors provide high-quality images, they are limited by weather and lighting conditions. In contrast, SAR sensors can operate effectively under adverse conditions. This letter proposes a GAN-based SAR-to-optical image translation method named Seg-CycleGAN, designed to enhance the accuracy of ship target translation by leveraging semantic information from a pre-trained semantic segmentation model. Our method utilizes the downstream task of ship target semantic segmentation to guide the training of image translation network, improving the quality of output Optical-styled images. The potential of foundation-model-annotated datasets in SAR-to-optical translation tasks is revealed. This work suggests broader research and applications for downstream-task-guided frameworks. The code will be available at https://github.com/NPULHH/, Comment: 8 pages, 5 figures

Traditional sperm morphology analysis is based on tedious manual annotation. Automated morphology analysis of a high number of sperm requires accurate segmentation of each sperm part and quantitative morphology evaluation. State-of-the-art instance-aware part segmentation networks follow a "detect-then-segment" paradigm. However, due to sperm's slim shape, their segmentation suffers from large context loss and feature distortion due to bounding box cropping and resizing during ROI Align. Moreover, morphology measurement of sperm tail is demanding because of the long and curved shape and its uneven width. This paper presents automated techniques to measure sperm morphology parameters automatically and quantitatively. A novel attention-based instance-aware part segmentation network is designed to reconstruct lost contexts outside bounding boxes and to fix distorted features, by refining preliminary segmented masks through merging features extracted by feature pyramid network. An automated centerline-based tail morphology measurement method is also proposed, in which an outlier filtering method and endpoint detection algorithm are designed to accurately reconstruct tail endpoints. Experimental results demonstrate that the proposed network outperformed the state-of-the-art top-down RP-R-CNN by 9.2% [AP]_vol^p, and the proposed automated tail morphology measurement method achieved high measurement accuracies of 95.34%,96.39%,91.2% for length, width and curvature, respectively., Comment: Accepted to ICRA 2024

Traversing 3-D complex environments has always been a significant challenge for legged locomotion. Existing methods typically rely on external sensors such as vision and lidar to preemptively react to obstacles by acquiring environmental information. However, in scenarios like nighttime or dense forests, external sensors often fail to function properly, necessitating robots to rely on proprioceptive sensors to perceive diverse obstacles in the environment and respond promptly. This task is undeniably challenging. Our research finds that methods based on collision detection can enhance a robot's perception of environmental obstacles. In this work, we propose an end-to-end learning-based quadruped robot motion controller that relies solely on proprioceptive sensing. This controller can accurately detect, localize, and agilely respond to collisions in unknown and complex 3D environments, thereby improving the robot's traversability in complex environments. We demonstrate in both simulation and real-world experiments that our method enables quadruped robots to successfully traverse challenging obstacles in various complex environments., Comment: 10 pages, 8 figures,submitted to iros2024

Porting codes to GPU often requires major efforts. While several tools exist for automatically offload numerical libraries such as BLAS and LAPACK, they often prove impractical due to the high cost of mandatory data transfer. The new unified memory architecture in NVIDIA Grace-Hopper allows high bandwidth cache-coherent memory access of all memory from both CPU and GPU, potentially eliminating bottleneck faced in conventional architecture. This breakthrough opens up new avenues for application development and porting strategies. In this study, we introduce a new tool for automatic BLAS offload, the tool leverages the high speed cache coherent NVLink C2C interconnect in Grace-Hopper, and enables performant GPU offload for BLAS heavy applications with no code changes or recompilation. The tool was tested on two quantum chemistry or physics codes, great performance benefits were observed.

In the field of brain science, data sharing across servers is becoming increasingly challenging due to issues such as industry competition, privacy security, and administrative procedure policies and regulations. Therefore, there is an urgent need to develop new methods for data analysis and processing that enable scientific collaboration without data sharing. In view of this, this study proposes to study and develop a series of efficient non-negative coupled tensor decomposition algorithm frameworks based on federated learning called FCNCP for the EEG data arranged on different servers. It combining the good discriminative performance of tensor decomposition in high-dimensional data representation and decomposition, the advantages of coupled tensor decomposition in cross-sample tensor data analysis, and the features of federated learning for joint modelling in distributed servers. The algorithm utilises federation learning to establish coupling constraints for data distributed across different servers. In the experiments, firstly, simulation experiments are carried out using simulated data, and stable and consistent decomposition results are obtained, which verify the effectiveness of the proposed algorithms in this study. Then the FCNCP algorithm was utilised to decompose the fifth-order event-related potential (ERP) tensor data collected by applying proprioceptive stimuli on the left and right hands. It was found that contralateral stimulation induced more symmetrical components in the activation areas of the left and right hemispheres. The conclusions drawn are consistent with the interpretations of related studies in cognitive neuroscience, demonstrating that the method can efficiently process higher-order EEG data and that some key hidden information can be preserved.

Microarchitecture simulators are indispensable tools for microarchitecture designers to validate, estimate, and optimize new hardware that meets specific design requirements. While the quest for a fast, accurate and detailed microarchitecture simulation has been ongoing for decades, existing simulators excel and fall short at different aspects: (i) Although execution-driven simulation is accurate and detailed, it is extremely slow and requires expert-level experience to design. (ii) Trace-driven simulation reuses the execution traces in pursuit of fast simulation but faces accuracy concerns and fails to achieve significant speedup. (iii) Emerging deep learning (DL)-based simulations are remarkably fast and have acceptable accuracy but fail to provide adequate low-level microarchitectural performance metrics crucial for microarchitectural bottleneck analysis. Additionally, they introduce substantial overheads from trace regeneration and model re-training when simulating a new microarchitecture. Re-thinking the advantages and limitations of the aforementioned simulation paradigms, this paper introduces TAO that redesigns the DL-based simulation with three primary contributions: First, we propose a new training dataset design such that the subsequent simulation only needs functional trace as inputs, which can be rapidly generated and reused across microarchitectures. Second, we redesign the input features and the DL model using self-attention to support predicting various performance metrics. Third, we propose techniques to train a microarchitecture agnostic embedding layer that enables fast transfer learning between different microarchitectural configurations and reduces the re-training overhead of conventional DL-based simulators. Our extensive evaluation shows TAO can reduce the overall training and simulation time by 18.06x over the state-of-the-art DL-based endeavors., Comment: Published in POMACS and SIGMETRICS'24

It was recently found that the optical field fluctuations in self-defocusing media can be described by sound waves propagating in a two-dimensional photon-fluid. This photon-fluid is controlled by the driving beam and serves as the background in which the sound waves can experience an effective curved spacetime, such that it provides a new platform of studying analog black holes. In this paper, we are interested in investigating the quasinormal modes (QNMs) of this analog black hole in the photon-fluid model. Based on the master equation of motion of the optical field fluctuations, we calculate the frequencies of quasinormal modes (QNF) with three different numerical methods to make sure the QNF we get are reliable. Besides fundamental modes, we also try to calculate the overtones up to $n=3$ aiming to uncover more properties of QNF. The effects of angular velocity $\Omega_H$ of the black hole, the overtone number $n$ and the winding number $m$ on the QNF are investigated. Under the $m$ with opposite sign, we find that both the real and imaginary part of the QNF will show strikingly contrasting behaviors when the QNF is plotted against $\Omega_H$, and the similar contrast effects are also found when comparing the influences from winding number and overtone number. We hope that this work may potentially contribute to the future detections of QNMs in experimental settings of photon-fluid., Comment: 32pages,6figures. Accepted for publication in PRD

Resistive Random Access Memory (ReRAM) has emerged as a promising platform for deep neural networks (DNNs) due to its support for parallel in-situ matrix-vector multiplication. However, hardware failures, such as stuck-at-fault defects, can result in significant prediction errors during model inference. While additional crossbars can be used to address these failures, they come with storage overhead and are not efficient in terms of space, energy, and cost. In this paper, we propose a fault protection mechanism that incurs zero space cost. Our approach includes: 1) differentiable structure pruning of rows and columns to reduce model redundancy, 2) weight duplication and voting for robust output, and 3) embedding duplicated most significant bits (MSBs) into the model weight. We evaluate our method on nine tasks of the GLUE benchmark with the BERT model, and experimental results prove its effectiveness.

We give two constructions of families of elliptic curves over cubic or quartic fields with three, respectively four, `integral' elements in the kernel of the tame symbol on the curves. The fields are in general non-Abelian, and the elements linearly independent. For their integrality, we prove a new criterion that does not ignore any torsion. We also verify Beilinson's conjecture numerically for just over 90 of the curves.

Chiral exact flat bands (FBs) at charge neutrality have attracted much recent interest, presenting an intriguing condensed-matter system to realize exact many-body phenomena, as specifically shown in "magic angle" twisted bilayer graphene for superconductivity and triangulene-based superatomic graphene for excitonic condensation. Yet, no generic physical model to realize such FBs has been developed. Here we present a new mathematical theorem, called bipartite double cover (BDC) theorem, and prove that the BDC of line-graph (LG) lattices hosts at least two chiral exact FBs of opposite chirality, i.e., yin-yang FBs, centered-around/at charge neutrality (E = 0) akin to the "chiral limit" of twisted bilayer graphene. We illustrate this theorem by mapping it exactly onto tight-binding lattice models of the BDC of LGs of hexagonal lattice for strong topological and of triangular lattice for fragile topological FBs, respectively. Moreover, we use orbital design principle to realize such exotic yin-yang FBs in non-BDC lattices to instigate their real material discovery. This work not only enables the search for exact chiral FBs at zero energy beyond moir\'e heterostructures, but also opens the door to discovering quantum semiconductor features with FB-enabled strongly correlated carriers.

Motivated by the need for analysing large spatio-temporal panel data, we introduce a novel dimensionality reduction methodology for $n$-dimensional random fields observed across a number $S$ spatial locations and $T$ time periods. We call it General Spatio-Temporal Factor Model (GSTFM). First, we provide the probabilistic and mathematical underpinning needed for the representation of a random field as the sum of two components: the common component (driven by a small number $q$ of latent factors) and the idiosyncratic component (mildly cross-correlated). We show that the two components are identified as $n\to\infty$. Second, we propose an estimator of the common component and derive its statistical guarantees (consistency and rate of convergence) as $\min(n, S, T )\to\infty$. Third, we propose an information criterion to determine the number of factors. Estimation makes use of Fourier analysis in the frequency domain and thus we fully exploit the information on the spatio-temporal covariance structure of the whole panel. Synthetic data examples illustrate the applicability of GSTFM and its advantages over the extant generalized dynamic factor model that ignores the spatial correlations.

Recent research highlights frequent model communication as a significant bottleneck to the efficiency of decentralized machine learning (ML), especially for large-scale and over-parameterized neural networks (NNs). To address this, we present Malcom-PSGD, a novel decentralized ML algorithm that combines gradient compression techniques with model sparsification. We promote model sparsity by adding $\ell_1$ regularization to the objective and present a decentralized proximal SGD method for training. Our approach employs vector source coding and dithering-based quantization for the compressed gradient communication of sparsified models. Our analysis demonstrates that Malcom-PSGD achieves a convergence rate of $\mathcal{O}(1/\sqrt{t})$ with respect to the iterations $t$, assuming a constant consensus and learning rate. This result is supported by our proof for the convergence of non-convex compressed Proximal SGD methods. Additionally, we conduct a bit analysis, providing a closed-form expression for the communication costs associated with Malcom-PSGD. Numerical results verify our theoretical findings and demonstrate that our method reduces communication costs by approximately $75\%$ when compared to the state-of-the-art.

With the rapid development of the internet, online social media welcomes people with different backgrounds through its diverse content. The increasing usage of emoji becomes a noticeable trend thanks to emoji's rich information beyond cultural or linguistic borders. However, the current study on emojis is limited to single emoji prediction and there are limited data resources available for further study of the interesting linguistic phenomenon. To this end, we synthesize a large text-emoji parallel corpus, Text2Emoji, from a large language model. Based on the parallel corpus, we distill a sequence-to-sequence model, EmojiLM, which is specialized in the text-emoji bidirectional translation. Extensive experiments on public benchmarks and human evaluation demonstrate that our proposed model outperforms strong baselines and the parallel corpus benefits emoji-related downstream tasks.

In the upcoming decade, deep learning may revolutionize the natural sciences, enhancing our capacity to model and predict natural occurrences. This could herald a new era of scientific exploration, bringing significant advancements across sectors from drug development to renewable energy. To answer this call, we present DeepSpeed4Science initiative (deepspeed4science.ai) which aims to build unique capabilities through AI system technology innovations to help domain experts to unlock today's biggest science mysteries. By leveraging DeepSpeed's current technology pillars (training, inference and compression) as base technology enablers, DeepSpeed4Science will create a new set of AI system technologies tailored for accelerating scientific discoveries by addressing their unique complexity beyond the common technical approaches used for accelerating generic large language models (LLMs). In this paper, we showcase the early progress we made with DeepSpeed4Science in addressing two of the critical system challenges in structural biology research.

In this paper, we propose a source coding scheme that represents data from unknown distributions through frequency and support information. Existing encoding schemes often compress data by sacrificing computational efficiency or by assuming the data follows a known distribution. We take advantage of the structure that arises within the spatial representation and utilize it to encode run-lengths within this representation using Golomb coding. Through theoretical analysis, we show that our scheme yields an overall bit rate that nears entropy without a computationally complex encoding algorithm and verify these results through numerical experiments., Comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

We develop an improved method to explore the $\Xi_c- \Xi_c'$ mixing which arises from the flavor SU(3) and heavy quark symmetry breaking. In this method, the flavor eigenstates under the SU(3) symmetry are at first constructed and the corresponding masses can be nonperturbatively determined. Matrix elements of the mass operators which break the flavor SU(3) symmetry sandwiched by the flavor eigenstates are then calculated. Diagonalizing the corresponding matrix of Hamiltonian gives the mass eigenstates of the full Hamiltonian and determines the mixing. Following the previous lattice QCD calculation of $\Xi_c$ and $\Xi_c'$, and estimating an off-diagonal matrix element, we extract the mixing angle between the $\Xi_c$ and $\Xi_c'$. Preliminary numerical results for the mixing angle confirm the previous observation that such mixing is incapable to explain the large SU(3) symmetry breaking in semileptonic decays of charmed baryons.

The MuST package is a computational software designed for ab initio electronic structure calculations for solids. The Locally Self-consistent Multiple Scattering (LSMS) method implemented in MuST allows to perform the electronic structure calculation for systems with a large number of atoms per unit cell. For the LSMS method with muffin-tin potential approximation, the major computational challenge is the matrix inverse for the scattering matrix calculation, which could take more than 90\% of the computing time. However, the matrix inverse can be significantly accelerated by modern graphical-processing-units (GPUs). In this paper, we discuss our approach to the code acceleration by offloading the matrix inverse tasks to the GPUs through a Fortran-C interface from the Fortran code to the CUDA code. We report our performance results showing significant speedup ratio achieved to the calculations of NiAu alloy, a candidate for thermoelectric materials.

Graph Neural Networks (GNNs) are becoming increasingly popular due to their superior performance in critical graph-related tasks. While quantization is widely used to accelerate GNN computation, quantized training faces unprecedented challenges. Current quantized GNN training systems often have longer training times than their full-precision counterparts for two reasons: (i) addressing the accuracy challenge leads to excessive overhead, and (ii) the optimization potential exposed by quantization is not adequately leveraged. This paper introduces Tango which re-thinks quantization challenges and opportunities for graph neural network training on GPUs with three contributions: Firstly, we introduce efficient rules to maintain accuracy during quantized GNN training. Secondly, we design and implement quantization-aware primitives and inter-primitive optimizations that can speed up GNN training. Finally, we integrate Tango with the popular Deep Graph Library (DGL) system and demonstrate its superior performance over state-of-the-art approaches on various GNN models and datasets.