arXiv subjects

Yang Wang

Publications and source records attributed to Yang Wang.

At least 487 records · Page 27Linked to original sources

arXiv2023

LUT-NN: Empower Efficient Neural Network Inference with Centroid Learning and Table Lookup

On-device Deep Neural Network (DNN) inference consumes significant computing resources and development efforts. To alleviate that, we propose LUT-NN, the first system to empower inference by table lookup, to reduce inference cost. LUT-NN learns the typical features for each operator, named centroid, and precompute the results for these centroids to save in lookup tables. During inference, the results of the closest centroids with the inputs can be read directly from the table, as the approximated outputs without computations. LUT-NN integrates two major novel techniques: (1) differentiable centroid learning through backpropagation, which adapts three levels of approximation to minimize the accuracy impact by centroids; (2) table lookup inference execution, which comprehensively considers different levels of parallelism, memory access reduction, and dedicated hardware units for optimal performance. LUT-NN is evaluated on multiple real tasks, covering image and speech recognition, and nature language processing. Compared to related work, LUT-NN improves accuracy by 66% to 92%, achieving similar level with the original models. LUT-NN reduces the cost at all dimensions, including FLOPs ($\leq$ 16x), model size ($\leq$ 7x), latency ($\leq$ 6.8x), memory ($\leq$ 6.5x), and power ($\leq$ 41.7%).

cs.LG↗

arXiv2023

Point-TTA: Test-Time Adaptation for Point Cloud Registration Using Multitask Meta-Auxiliary Learning

We present Point-TTA, a novel test-time adaptation framework for point cloud registration (PCR) that improves the generalization and the performance of registration models. While learning-based approaches have achieved impressive progress, generalization to unknown testing environments remains a major challenge due to the variations in 3D scans. Existing methods typically train a generic model and the same trained model is applied on each instance during testing. This could be sub-optimal since it is difficult for the same model to handle all the variations during testing. In this paper, we propose a test-time adaptation approach for PCR. Our model can adapt to unseen distributions at test-time without requiring any prior knowledge of the test data. Concretely, we design three self-supervised auxiliary tasks that are optimized jointly with the primary PCR task. Given a test instance, we adapt our model using these auxiliary tasks and the updated model is used to perform the inference. During training, our model is trained using a meta-auxiliary learning approach, such that the adapted model via auxiliary tasks improves the accuracy of the primary task. Experimental results demonstrate the effectiveness of our approach in improving generalization of point cloud registration and outperforming other state-of-the-art approaches.

cs.CV↗

arXiv2023

Test-Time Adaptation for Point Cloud Upsampling Using Meta-Learning

Affordable 3D scanners often produce sparse and non-uniform point clouds that negatively impact downstream applications in robotic systems. While existing point cloud upsampling architectures have demonstrated promising results on standard benchmarks, they tend to experience significant performance drops when the test data have different distributions from the training data. To address this issue, this paper proposes a test-time adaption approach to enhance model generality of point cloud upsampling. The proposed approach leverages meta-learning to explicitly learn network parameters for test-time adaption. Our method does not require any prior information about the test data. During meta-training, the model parameters are learned from a collection of instance-level tasks, each of which consists of a sparse-dense pair of point clouds from the training data. During meta-testing, the trained model is fine-tuned with a few gradient updates to produce a unique set of network parameters for each test instance. The updated model is then used for the final prediction. Our framework is generic and can be applied in a plug-and-play manner with existing backbone networks in point cloud upsampling. Extensive experiments demonstrate that our approach improves the performance of state-of-the-art models.

cs.CV↗

arXiv2023

High Performance GPU Accelerated MuST Software

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.

physics.comp-ph↗

arXiv2023

Practice of Alibaba Cloud on Elastic Resource Provisioning for Large-scale Microservices Cluster

Cloud-native architecture is becoming increasingly crucial for today's cloud computing environments due to the need for speed and flexibility in developing applications. It utilizes microservice technology to break down traditional monolithic applications into light-weight and self-contained microservice components. However, as microservices grow in scale and have dynamic inter-dependencies, they also pose new challenges in resource provisioning that cannot be fully addressed by traditional resource scheduling approaches. The various microservices with different resource needs and latency requirements can create complex calling chains, making it difficult to provide fine-grained and accurate resource allocation to each component while maintaining the overall quality of service in the chain. In this work, we aim to address the research problem on how to efficiently provision resources for the growing scale of microservice platform and ensure the performance of latency-critical microservices. To address the problem, we present in-depth analyses of Alibaba's microservice cluster and propose optimized resource provisioning algorithms to enhance resource utilization while ensuring the latency requirement. First, we analyze the distinct features of microservices in Alibaba's cluster compared to traditional applications. Then we present Alibaba's resource capacity provisioning workflow and framework to address challenges in resource provisioning for large-scale and latency-critical microservice clusters. Finally, we propose enhanced resource provisioning algorithms over Alibaba's current practice by making both proactive and reactive scheduling decisions based on different workloads patterns, which can improve resource usage by 10-15% in Alibaba's clusters, while maintaining the necessary latency for microservices.

cs.DC↗

arXiv2023

Building explainable graph neural network by sparse learning for the drug-protein binding prediction

Explainable Graph Neural Networks (GNNs) have been developed and applied to drug-protein binding prediction to identify the key chemical structures in a drug that have active interactions with the target proteins. However, the key structures identified by the current explainable GNN models are typically chemically invalid. Furthermore, a threshold needs to be manually selected to pinpoint the key structures from the rest. To overcome the limitations of the current explainable GNN models, we propose our SLGNN, which stands for using Sparse Learning to Graph Neural Networks. Our SLGNN relies on using a chemical-substructure-based graph (where nodes are chemical substructures) to represent a drug molecule. Furthermore, SLGNN incorporates generalized fussed lasso with message-passing algorithms to identify connected subgraphs that are critical for the drug-protein binding prediction. Due to the use of the chemical-substructure-based graph, it is guaranteed that any subgraphs in a drug identified by our SLGNN are chemically valid structures. These structures can be further interpreted as the key chemical structures for the drug to bind to the target protein. We demonstrate the explanatory power of our SLGNN by first showing all the key structures identified by our SLGNN are chemically valid. In addition, we illustrate that the key structures identified by our SLGNN have more predictive power than the key structures identified by the competing methods. At last, we use known drug-protein binding data to show the key structures identified by our SLGNN contain most of the binding sites.

q-bio.BM↗

arXiv2023

Replace Scoring with Arrangement: A Contextual Set-to-Arrangement Framework for Learning-to-Rank

Learning-to-rank is a core technique in the top-N recommendation task, where an ideal ranker would be a mapping from an item set to an arrangement (a.k.a. permutation). Most existing solutions fall in the paradigm of probabilistic ranking principle (PRP), i.e., first score each item in the candidate set and then perform a sort operation to generate the top ranking list. However, these approaches neglect the contextual dependence among candidate items during individual scoring, and the sort operation is non-differentiable. To bypass the above issues, we propose Set-To-Arrangement Ranking (STARank), a new framework directly generates the permutations of the candidate items without the need for individually scoring and sort operations; and is end-to-end differentiable. As a result, STARank can operate when only the ground-truth permutations are accessible without requiring access to the ground-truth relevance scores for items. For this purpose, STARank first reads the candidate items in the context of the user browsing history, whose representations are fed into a Plackett-Luce module to arrange the given items into a list. To effectively utilize the given ground-truth permutations for supervising STARank, we leverage the internal consistency property of Plackett-Luce models to derive a computationally efficient list-wise loss. Experimental comparisons against 9 the state-of-the-art methods on 2 learning-to-rank benchmark datasets and 3 top-N real-world recommendation datasets demonstrate the superiority of STARank in terms of conventional ranking metrics. Notice that these ranking metrics do not consider the effects of the contextual dependence among the items in the list, we design a new family of simulation-based ranking metrics, where existing metrics can be regarded as special cases. STARank can consistently achieve better performance in terms of PBM and UBM simulation-based metrics.

cs.IR↗

arXiv2023

Efficient Learned Lossless JPEG Recompression

JPEG is one of the most popular image compression methods. It is beneficial to compress those existing JPEG files without introducing additional distortion. In this paper, we propose a deep learning based method to further compress JPEG images losslessly. Specifically, we propose a Multi-Level Parallel Conditional Modeling (ML-PCM) architecture, which enables parallel decoding in different granularities. First, luma and chroma are processed independently to allow parallel coding. Second, we propose pipeline parallel context model (PPCM) and compressed checkerboard context model (CCCM) for the effective conditional modeling and efficient decoding within luma and chroma components. Our method has much lower latency while achieves better compression ratio compared with previous SOTA. After proper software optimization, we can obtain a good throughput of 57 FPS for 1080P images on NVIDIA T4 GPU. Furthermore, combined with quantization, our approach can also act as a lossy JPEG codec which has obvious advantage over SOTA lossy compression methods in high bit rate (bpp$>0.9$).

eess.IV↗

arXiv2023

Attention-Based Acoustic Feature Fusion Network for Depression Detection

Depression, a common mental disorder, significantly influences individuals and imposes considerable societal impacts. The complexity and heterogeneity of the disorder necessitate prompt and effective detection, which nonetheless, poses a difficult challenge. This situation highlights an urgent requirement for improved detection methods. Exploiting auditory data through advanced machine learning paradigms presents promising research directions. Yet, existing techniques mainly rely on single-dimensional feature models, potentially neglecting the abundance of information hidden in various speech characteristics. To rectify this, we present the novel Attention-Based Acoustic Feature Fusion Network (ABAFnet) for depression detection. ABAFnet combines four different acoustic features into a comprehensive deep learning model, thereby effectively integrating and blending multi-tiered features. We present a novel weight adjustment module for late fusion that boosts performance by efficaciously synthesizing these features. The effectiveness of our approach is confirmed via extensive validation on two clinical speech databases, CNRAC and CS-NRAC, thereby outperforming previous methods in depression detection and subtype classification. Further in-depth analysis confirms the key role of each feature and highlights the importance of MFCCrelated features in speech-based depression detection.

cs.SD↗

arXiv2023

The Snowflake Hypothesis: Training Deep GNN with One Node One Receptive field

Despite Graph Neural Networks demonstrating considerable promise in graph representation learning tasks, GNNs predominantly face significant issues with over-fitting and over-smoothing as they go deeper as models of computer vision realm. In this work, we conduct a systematic study of deeper GNN research trajectories. Our findings indicate that the current success of deep GNNs primarily stems from (I) the adoption of innovations from CNNs, such as residual/skip connections, or (II) the tailor-made aggregation algorithms like DropEdge. However, these algorithms often lack intrinsic interpretability and indiscriminately treat all nodes within a given layer in a similar manner, thereby failing to capture the nuanced differences among various nodes. To this end, we introduce the Snowflake Hypothesis -- a novel paradigm underpinning the concept of ``one node, one receptive field''. The hypothesis draws inspiration from the unique and individualistic patterns of each snowflake, proposing a corresponding uniqueness in the receptive fields of nodes in the GNNs. We employ the simplest gradient and node-level cosine distance as guiding principles to regulate the aggregation depth for each node, and conduct comprehensive experiments including: (1) different training schemes; (2) various shallow and deep GNN backbones, and (3) various numbers of layers (8, 16, 32, 64) on multiple benchmarks (six graphs including dense graphs with millions of nodes); (4) compare with different aggregation strategies. The observational results demonstrate that our hypothesis can serve as a universal operator for a range of tasks, and it displays tremendous potential on deep GNNs. It can be applied to various GNN frameworks, enhancing its effectiveness when operating in-depth, and guiding the selection of the optimal network depth in an explainable and generalizable way.

cs.LG↗

arXiv2023

ARAI-MVSNet: A multi-view stereo depth estimation network with adaptive depth range and depth interval

Multi-View Stereo~(MVS) is a fundamental problem in geometric computer vision which aims to reconstruct a scene using multi-view images with known camera parameters. However, the mainstream approaches represent the scene with a fixed all-pixel depth range and equal depth interval partition, which will result in inadequate utilization of depth planes and imprecise depth estimation. In this paper, we present a novel multi-stage coarse-to-fine framework to achieve adaptive all-pixel depth range and depth interval. We predict a coarse depth map in the first stage, then an Adaptive Depth Range Prediction module is proposed in the second stage to zoom in the scene by leveraging the reference image and the obtained depth map in the first stage and predict a more accurate all-pixel depth range for the following stages. In the third and fourth stages, we propose an Adaptive Depth Interval Adjustment module to achieve adaptive variable interval partition for pixel-wise depth range. The depth interval distribution in this module is normalized by Z-score, which can allocate dense depth hypothesis planes around the potential ground truth depth value and vice versa to achieve more accurate depth estimation. Extensive experiments on four widely used benchmark datasets~(DTU, TnT, BlendedMVS, ETH 3D) demonstrate that our model achieves state-of-the-art performance and yields competitive generalization ability. Particularly, our method achieves the highest Acc and Overall on the DTU dataset, while attaining the highest Recall and $F_{1}$-score on the Tanks and Temples intermediate and advanced dataset. Moreover, our method also achieves the lowest $e_{1}$ and $e_{3}$ on the BlendedMVS dataset and the highest Acc and $F_{1}$-score on the ETH 3D dataset, surpassing all listed methods.Project website: https://github.com/zs670980918/ARAI-MVSNet

cs.CV↗

arXiv2023

Production and chemical separation of $^{229}$Pa toward observation of $γ$ rays of $^{229\rm{m}}$Th

To observe the $γ$ rays emitted from the low-lying isomeric state of $^{229}$Th ($^{229\rm{m}}$Th), we aim to dope fluoride crystals with its precursor $^{229}$Pa. In this study, we produced $^{229}$Pa by a 30-MeV proton bombardment on $^{232}$Th and developed a chemical separation method. The chemical yield of Pa was 93(4)%, and the physical production yield of $^{229}$Pa for the proton energy range of 29.0${-}$30.1 MeV was measured to be 9.4(8) MBq/${\rmμ}$Ah, which was more than 10 times higher than those of $^{232,230,228,233}$Pa. These high chemical and production yields will allow us to prepare fluoride crystals having a sufficient amount of $^{229}$Pa for the observation of the $γ$ rays of $^{229\rm{m}}$Th.

physics.ins-det↗

arXiv2023

ARPES Detection of Superconducting Gap Sign in Unconventional Superconductors

Superconductivity is realized by opening a gap in the superconducting state. The gap symmetry is crucial in understanding the underlying superconductivity mechanism. The magnitude and the phase are essential in fully characterizing the superconducting gap. Angle-resolved photoemission spectroscopy (ARPES) has played a key role in determining the gap symmetry in unconventional superconductors. However, it has been considered so far that ARPES can only measure the magnitude of the superconducting gap but not its phase; the phase has to be detected by other phase-sensitive techniques. Here we propose a new method to directly detect the superconducting gap sign by using ARPES. This method is successfully validated in a cuprate superconductor with a well-known $d$-wave gap symmetry. When two bands are nearby in momentum space and have a strong interband interaction, the resulted electronic structures in the superconducting state are sensitive to the relative gap sign between the two bands which can be captured by ARPES measurements. Our present work provides a new way to detect the gap sign and can be applied to various superconductors, particularly those with multiple orbitals like the iron-based superconductors. It also makes ARPES more powerful to determine both the gap magnitude and the phase that are significant in understanding the superconductivity mechanism of unconventional superconductors.

cond-mat.supr-con↗

arXiv2023

Hydrophobic Silica Microcavities with Sustainable Nonlinear Photonic Performance

Ultrahigh quality factor (Q) microcavities have been emerging as an appealing compact photonic platform for various applications. The Q factor plays a critical role in determining the nonlinear optical performance of a microcavity. However, a silica microcavity suffers from severe degradation of its Q value over time during storage or use in air due to the accumulating surface absorption loss, which would deteriorate their nonlinear photonic performance. Here, we report a new type of ultrahigh Q silica microcavity that effectively prevents the Q degradation over time. The Q values of the devices remain unchanged over time under storage in air. Optical frequency combs are generated with sustainable ultralow threshold performance in the course of time from the devices in open air. This approach would greatly facilitate ultrahigh Q silica-based photonic devices for next generation photonic applications.

physics.optics↗

arXiv2023

Design of the offline test electronics for the nozzle system of proton therapy

A set of nozzle equipment for proton therapy is now being developed at China Institute of Atomic Energy. To facilitate the off-line commissioning of the whole equipment, a set of ionization chamber signal generation system, the test electronics, is designed. The system uses ZYNQ SoC as the main control unit and outputs the beam dose analog signal through DAC8532. The dual SPDT analog switch, DG636, is used to simulate the beam position signals according to Gaussian distribution. The results show that the system can simulate the beam position, dose, and other related analog signals generated by the proton beam when passing through the ionization chamber. Moreover, the accuracy of the simulated beam position is within +/-0.33mm, and the accuracy of the simulated dose signal is within +/-1%. At the same time, it can output analog signals representing environmental parameters. The test electronics meets the design requirements, which can be used to commission the nozzle system as well as the treatment control system without the proton beam.

physics.acc-ph↗

arXiv2023

Nonlinear current response of two-dimensional systems under in-plane magnetic field

We theoretically investigate the nonlinear response current of a two-dimensional system under an in-plane magnetic field. Based on the extended semiclassical theory, we develop a unified theory including both longitudinal and transverse currents and classify contributions according to their scaling with the relaxation time. Besides time-reversal-even contributions, we reveal a previously unknown time-reversal-odd contribution to the Hall current, which occurs in magnetic systems, exhibits band geometric origin, and is linear in relaxation time. We show that the different contributions exhibit different symmetry characters, especially in their angular dependence on the field orientation, which can be used to distinguish them in experiment. The theory is explicitly demonstrated in the study of the Rashba model. Our work presents a deepened understanding of nonlinear planar transport, proposes approaches to distinguish different contributions, and sheds light on possible routes to enhance the effect in practice.

cond-mat.mes-hall↗

arXiv2023

Understanding Privacy Over-collection in WeChat Sub-app Ecosystem

Nowadays the app-in-app paradigm is becoming increasingly popular, and sub-apps have become an important form of mobile applications. WeChat, the leading app-in-app platform, provides millions of sub-apps that can be used for online shopping, financing, social networking, etc. However, privacy issues in this new ecosystem have not been well understood. This paper performs the first systematic study of privacy over-collection in sub-apps (denoted as SPO), where sub-apps actually collect more privacy data than they claim in their privacy policies. We propose a taxonomy of privacy for this ecosystem and a framework named SPOChecker to automatically detect SPO in real-world sub-apps. Based on SPOChecker, we collect 5,521 popular and representative WeChat sub-apps and conduct a measurement study to understand SPO from three aspects: its landscape, accountability, and defense methods. The result is worrisome, that more than half of all studied sub-apps do not provide users with privacy policies. Among 2,511 sub-apps that provide privacy policies, 489 (19.47%) of them contain SPO. We look into the detailed characteristics of SPO, figure out possible reasons and the responsibilities of stakeholders in the ecosystem, and rethink current defense methods. The measurement leads to several insightful findings that can help the community to better understand SPO and protect privacy in sub-apps.

cs.CR↗

arXiv2023

DGraph: A Large-Scale Financial Dataset for Graph Anomaly Detection

Graph Anomaly Detection (GAD) has recently become a hot research spot due to its practicability and theoretical value. Since GAD emphasizes the application and the rarity of anomalous samples, enriching the varieties of its datasets is fundamental work. Thus, this paper present DGraph, a real-world dynamic graph in the finance domain. DGraph overcomes many limitations of current GAD datasets. It contains about 3M nodes, 4M dynamic edges, and 1M ground-truth nodes. We provide a comprehensive observation of DGraph, revealing that anomalous nodes and normal nodes generally have different structures, neighbor distribution, and temporal dynamics. Moreover, it suggests that unlabeled nodes are also essential for detecting fraudsters. Furthermore, we conduct extensive experiments on DGraph. Observation and experiments demonstrate that DGraph is propulsive to advance GAD research and enable in-depth exploration of anomalous nodes.

cs.SI↗