arXiv subjects

Lei Huang

Publications and source records attributed to Lei Huang.

At least 217 records · Page 12Linked to original sources

arXiv2020

Convolutional Neural Network Training with Distributed K-FAC

Training neural networks with many processors can reduce time-to-solution; however, it is challenging to maintain convergence and efficiency at large scales. The Kronecker-factored Approximate Curvature (K-FAC) was recently proposed as an approximation of the Fisher Information Matrix that can be used in natural gradient optimizers. We investigate here a scalable K-FAC design and its applicability in convolutional neural network (CNN) training at scale. We study optimization techniques such as layer-wise distribution strategies, inverse-free second-order gradient evaluation, and dynamic K-FAC update decoupling to reduce training time while preserving convergence. We use residual neural networks (ResNet) applied to the CIFAR-10 and ImageNet-1k datasets to evaluate the correctness and scalability of our K-FAC gradient preconditioner. With ResNet-50 on the ImageNet-1k dataset, our distributed K-FAC implementation converges to the 75.9% MLPerf baseline in 18-25% less time than does the classic stochastic gradient descent (SGD) optimizer across scales on a GPU cluster.

cs.LG↗

arXiv2020

The GWAC Data Processing and Management System

GWAC will have been built an integrated FOV of 5,000 $degree^2$ and have already built 1,800 square $degree^2$. The limit magnitude of a 10-second exposure image in the moonless night is 16R. In each observation night, GWAC produces about 0.7TB of raw data, and the data processing pipeline generates millions of single frame alerts. We describe the GWAC Data Processing and Management System (GPMS), including hardware architecture, database, detection-filtering-validation of transient candidates, data archiving, and user interfaces for the check of transient and the monitor of the system. GPMS combines general technology and software in astronomy and computer field, and use some advanced technologies such as deep learning. Practical results show that GPMS can fully meet the scientific data processing requirement of GWAC. It can online accomplish the detection, filtering and validation of millions of transient candidates, and feedback the final results to the astronomer in real-time. During the observation from October of 2018 to December of 2019, we have already found 102 transients.

astro-ph.IM↗

arXiv2020

Topic Detection and Summarization of User Reviews

A massive amount of reviews are generated daily from various platforms. It is impossible for people to read through tons of reviews and to obtain useful information. Automatic summarizing customer reviews thus is important for identifying and extracting the essential information to help users to obtain the gist of the data. However, as customer reviews are typically short, informal, and multifaceted, it is extremely challenging to generate topic-wise summarization.While there are several studies aims to solve this issue, they are heuristic methods that are developed only utilizing customer reviews. Unlike existing method, we propose an effective new summarization method by analyzing both reviews and summaries.To do that, we first segment reviews and summaries into individual sentiments. As the sentiments are typically short, we combine sentiments talking about the same aspect into a single document and apply topic modeling method to identify hidden topics among customer reviews and summaries. Sentiment analysis is employed to distinguish positive and negative opinions among each detected topic. A classifier is also introduced to distinguish the writing pattern of summaries and that of customer reviews. Finally, sentiments are selected to generate the summarization based on their topic relevance, sentiment analysis score and the writing pattern. To test our method, a new dataset comprising product reviews and summaries about 1028 products are collected from Amazon and CNET. Experimental results show the effectiveness of our method compared with other methods.

cs.CL↗

arXiv2020

Toward a Full MHD Jet Model of Spinning Black Holes--II: Kinematics and Application to the M87 Jet

In this paper, we investigate the magnetohydrodynamical structure of a jet powered by a spinning black hole, where electromagnetic fields and fluid motion are governed by the Grad-Shafranov equation and the Bernoulli equation, respectively. Assuming steady and axisymmetric jet structure, the global solution is uniquely determined with prescribed plasma loading into the jet and the poloidal shape of the outmost magnetic field line. We apply this model to the jet in the center of nearby radio galaxy M87, and we find it can naturally explain the slow flow acceleration and the flow velocity stratification within $10^5$ gravitational radii from the central black hole. In particular, we find the extremal black hole spin is disfavored by the flow velocity measurements, if the plasma loading to the jet is dominated by the electron/positron pair production at the jet base.

astro-ph.HE↗

arXiv2020

Controllable Orthogonalization in Training DNNs

Orthogonality is widely used for training deep neural networks (DNNs) due to its ability to maintain all singular values of the Jacobian close to 1 and reduce redundancy in representation. This paper proposes a computationally efficient and numerically stable orthogonalization method using Newton's iteration (ONI), to learn a layer-wise orthogonal weight matrix in DNNs. ONI works by iteratively stretching the singular values of a weight matrix towards 1. This property enables it to control the orthogonality of a weight matrix by its number of iterations. We show that our method improves the performance of image classification networks by effectively controlling the orthogonality to provide an optimal tradeoff between optimization benefits and representational capacity reduction. We also show that ONI stabilizes the training of generative adversarial networks (GANs) by maintaining the Lipschitz continuity of a network, similar to spectral normalization (SN), and further outperforms SN by providing controllable orthogonality.

cs.CV↗

arXiv2020

An Efficient Agreement Mechanism in CapsNets By Pairwise Product

Capsule networks (CapsNets) are capable of modeling visual hierarchical relationships, which is achieved by the "routing-by-agreement" mechanism. This paper proposes a pairwise agreement mechanism to build capsules, inspired by the feature interactions of factorization machines (FMs). The proposed method has a much lower computation complexity. We further proposed a new CapsNet architecture that combines the strengths of residual networks in representing low-level visual features and CapsNets in modeling the relationships of parts to wholes. We conduct comprehensive experiments to compare the routing algorithms, including dynamic routing, EM routing, and our proposed FM agreement, based on both architectures of original CapsNet and our proposed one, and the results show that our method achieves both excellent performance and efficiency under a variety of situations.

cs.CV↗

arXiv2020

An Investigation into the Stochasticity of Batch Whitening

Batch Normalization (BN) is extensively employed in various network architectures by performing standardization within mini-batches. A full understanding of the process has been a central target in the deep learning communities. Unlike existing works, which usually only analyze the standardization operation, this paper investigates the more general Batch Whitening (BW). Our work originates from the observation that while various whitening transformations equivalently improve the conditioning, they show significantly different behaviors in discriminative scenarios and training Generative Adversarial Networks (GANs). We attribute this phenomenon to the stochasticity that BW introduces. We quantitatively investigate the stochasticity of different whitening transformations and show that it correlates well with the optimization behaviors during training. We also investigate how stochasticity relates to the estimation of population statistics during inference. Based on our analysis, we provide a framework for designing and comparing BW algorithms in different scenarios. Our proposed BW algorithm improves the residual networks by a significant margin on ImageNet classification. Besides, we show that the stochasticity of BW can improve the GAN's performance with, however, the sacrifice of the training stability.

cs.CV↗

arXiv2020

The Medical Scribe: Corpus Development and Model Performance Analyses

There is a growing interest in creating tools to assist in clinical note generation using the audio of provider-patient encounters. Motivated by this goal and with the help of providers and medical scribes, we developed an annotation scheme to extract relevant clinical concepts. We used this annotation scheme to label a corpus of about 6k clinical encounters. This was used to train a state-of-the-art tagging model. We report ontologies, labeling results, model performances, and detailed analyses of the results. Our results show that the entities related to medications can be extracted with a relatively high accuracy of 0.90 F-score, followed by symptoms at 0.72 F-score, and conditions at 0.57 F-score. In our task, we not only identify where the symptoms are mentioned but also map them to canonical forms as they appear in the clinical notes. Of the different types of errors, in about 19-38% of the cases, we find that the model output was correct, and about 17-32% of the errors do not impact the clinical note. Taken together, the models developed in this work are more useful than the F-scores reflect, making it a promising approach for practical applications.

cs.CL↗

arXiv2020

Unbiased Scene Graph Generation via Rich and Fair Semantic Extraction

Extracting graph representation of visual scenes in image is a challenging task in computer vision. Although there has been encouraging progress of scene graph generation in the past decade, we surprisingly find that the performance of existing approaches is largely limited by the strong biases, which mainly stem from (1) unconsciously assuming relations with certain semantic properties such as symmetric and (2) imbalanced annotations over different relations. To alleviate the negative effects of these biases, we proposed a new and simple architecture named Rich and Fair semantic extraction network (RiFa for short), to not only capture rich semantic properties of the relations, but also fairly predict relations with different scale of annotations. Using pseudo-siamese networks, RiFa embeds the subject and object respectively to distinguish their semantic differences and meanwhile preserve their underlying semantic properties. Then, it further predicts subject-object relations based on both the visual and semantic features of entities under certain contextual area, and fairly ranks the relation predictions for those with a few annotations. Experiments on the popular Visual Genome dataset show that RiFa achieves state-of-the-art performance under several challenging settings of scene graph task. Especially, it performs significantly better on capturing different semantic properties of relations, and obtains the best overall per relation performance.

cs.CV↗

arXiv2019

Deformable Tube Network for Action Detection in Videos

We address the problem of spatio-temporal action detection in videos. Existing methods commonly either ignore temporal context in action recognition and localization, or lack the modelling of flexible shapes of action tubes. In this paper, we propose a two-stage action detector called Deformable Tube Network (DTN), which is composed of a Deformation Tube Proposal Network (DTPN) and a Deformable Tube Recognition Network (DTRN) similar to the Faster R-CNN architecture. In DTPN, a fast proposal linking algorithm (FTL) is introduced to connect region proposals across frames to generate multiple deformable action tube proposals. To perform action detection, we design a 3D convolution network with skip connections for tube classification and regression. Modelling action proposals as deformable tubes explicitly considers the shape of action tubes compared to 3D cuboids. Moreover, 3D convolution based recognition network can learn temporal dynamics sufficiently for action detection. Our experimental results show that we significantly outperform the methods with 3D cuboids and obtain the state-of-the-art results on both UCF-Sports and AVA datasets.

cs.CV↗

arXiv2019

Towards a Full MHD Jet Model of Spinning Black Holes--I: Framework and a split monopole example

In this paper, we construct a framework for investigating magnetohydrodynamical jet structure of spinning black holes (BHs), where electromagnetic fields and fluid motion are governed by the Grad-Shafranov equation and the Bernoulli equation, respectively. Assuming steady and axisymmetric jet structure, we can self-consistently obtain electromagnetic fields, fluid energy density and velocity within the jet, given proper plasma loading and boundary conditions. Specifically, we structure the two coupled governing equations as two eigenvalue problems, and develop full numerical techniques for solving them. As an example, we explicitly solve the governing equations for the split monopole magnetic field configuration and simplified plasma loading on the stagnation surface where the poloidal fluid velocity vanishes. As expected, we find the rotation of magnetic field lines is dragged down by fluid inertia, and the fluid as a whole does not contribute to energy extraction from the central BH, i.e., the magnetic Penrose process is not working. However, if we decompose the charged fluid as two oppositely charged components, we find the magnetic Penrose process does work for one of the two components when the plasma loading is low enough.

astro-ph.HE↗

arXiv2019

A Survey on Matrix Completion: Perspective of Signal Processing

Matrix completion (MC) is a promising technique which is able to recover an intact matrix with low-rank property from sub-sampled/incomplete data. Its application varies from computer vision, signal processing to wireless network, and thereby receives much attention in the past several years. There are plenty of works addressing the behaviors and applications of MC methodologies. This work provides a comprehensive review for MC approaches from the perspective of signal processing. In particular, the MC problem is first grouped into six optimization problems to help readers understand MC algorithms. Next, four representative types of optimization algorithms solving the MC problem are reviewed. Ultimately, three different application fields of MC are described and evaluated.

eess.SP↗

arXiv2019

Query-Adaptive Hash Code Ranking for Large-Scale Multi-View Visual Search

Hash based nearest neighbor search has become attractive in many applications. However, the quantization in hashing usually degenerates the discriminative power when using Hamming distance ranking. Besides, for large-scale visual search, existing hashing methods cannot directly support the efficient search over the data with multiple sources, and while the literature has shown that adaptively incorporating complementary information from diverse sources or views can significantly boost the search performance. To address the problems, this paper proposes a novel and generic approach to building multiple hash tables with multiple views and generating fine-grained ranking results at bitwise and tablewise levels. For each hash table, a query-adaptive bitwise weighting is introduced to alleviate the quantization loss by simultaneously exploiting the quality of hash functions and their complement for nearest neighbor search. From the tablewise aspect, multiple hash tables are built for different data views as a joint index, over which a query-specific rank fusion is proposed to rerank all results from the bitwise ranking by diffusing in a graph. Comprehensive experiments on image search over three well-known benchmarks show that the proposed method achieves up to 17.11% and 20.28% performance gains on single and multiple table search over state-of-the-art methods.

cs.IR↗

arXiv2019

Iterative Normalization: Beyond Standardization towards Efficient Whitening

Batch Normalization (BN) is ubiquitously employed for accelerating neural network training and improving the generalization capability by performing standardization within mini-batches. Decorrelated Batch Normalization (DBN) further boosts the above effectiveness by whitening. However, DBN relies heavily on either a large batch size, or eigen-decomposition that suffers from poor efficiency on GPUs. We propose Iterative Normalization (IterNorm), which employs Newton's iterations for much more efficient whitening, while simultaneously avoiding the eigen-decomposition. Furthermore, we develop a comprehensive study to show IterNorm has better trade-off between optimization and generalization, with theoretical and experimental support. To this end, we exclusively introduce Stochastic Normalization Disturbance (SND), which measures the inherent stochastic uncertainty of samples when applied to normalization operations. With the support of SND, we provide natural explanations to several phenomena from the perspective of optimization, e.g., why group-wise whitening of DBN generally outperforms full-whitening and why the accuracy of BN degenerates with reduced batch sizes. We demonstrate the consistently improved performance of IterNorm with extensive experiments on CIFAR-10 and ImageNet over BN and DBN.

cs.CV↗

arXiv2019

Temporal phase unwrapping using deep learning

The multi-frequency temporal phase unwrapping (MF-TPU) method, as a classical phase unwrapping algorithm for fringe projection profilometry (FPP), is capable of eliminating the phase ambiguities even in the presence of surface discontinuities or spatially isolated objects. For the simplest and most efficient case, two sets of 3-step phase-shifting fringe patterns are used: the high-frequency one is for 3D measurement and the unit-frequency one is for unwrapping the phase obtained from the high-frequency pattern set. The final measurement precision or sensitivity is determined by the number of fringes used within the high-frequency pattern, under the precondition that the phase can be successfully unwrapped without triggering the fringe order error. Consequently, in order to guarantee a reasonable unwrapping success rate, the fringe number (or period number) of the high-frequency fringe patterns is generally restricted to about 16, resulting in limited measurement accuracy. On the other hand, using additional intermediate sets of fringe patterns can unwrap the phase with higher frequency, but at the expense of a prolonged pattern sequence. Inspired by recent successes of deep learning techniques for computer vision and computational imaging, in this work, we report that the deep neural networks can learn to perform TPU after appropriate training, as called deep-learning based temporal phase unwrapping (DL-TPU), which can substantially improve the unwrapping reliability compared with MF-TPU even in the presence of different types of error sources, e.g., intensity noise, low fringe modulation, and projector nonlinearity. We further experimentally demonstrate for the first time, to our knowledge, that the high-frequency phase obtained from 64-period 3-step phase-shifting fringe patterns can be directly and reliably unwrapped from one unit-frequency phase using DL-TPU.

eess.IV↗

arXiv2019

Utilizing Complex-valued Network for Learning to Compare Image Patches

At present, the great achievements of convolutional neural network(CNN) in feature and metric learning have attracted many researchers. However, the vast majority of deep network architectures have been used to represent based on real values. The research of complex-valued networks is seldom concerned due to the absence of effective models and suitable distance of complex-valued vector. Motived by recent works, complex vectors have been shown to have a richer representational capacity and efficient complex blocks have been reported, we propose a new approach for learning image descriptors with complex numbers to compare image patches. We also propose a new architecture to learn image similarity function directly based on complex-valued network. We show that our models can perform competitive results on benchmark datasets. We make the source code of our models publicly available.

cs.CV↗

arXiv2019

Analyzing Periodicity and Saliency for Adult Video Detection

Content-based adult video detection plays an important role in preventing pornography. However, existing methods usually rely on single modality and seldom focus on multi-modality semantics representation. Addressing at this problem, we put forward an approach of analyzing periodicity and saliency for adult video detection. At first, periodic patterns and salient regions are respective-ly analyzed in audio-frames and visual-frames. Next, the multi-modal co-occurrence semantics is described by combining audio periodicity with visual saliency. Moreover, the performance of our approach is evaluated step by step. Experimental results show that our approach obviously outper-forms some state-of-the-art methods.

cs.CV↗

arXiv2018

FanStore: Enabling Efficient and Scalable I/O for Distributed Deep Learning

Emerging Deep Learning (DL) applications introduce heavy I/O workloads on computer clusters. The inherent long lasting, repeated, and random file access pattern can easily saturate the metadata and data service and negatively impact other users. In this paper, we present FanStore, a transient runtime file system that optimizes DL I/O on existing hardware/software stacks. FanStore distributes datasets to the local storage of compute nodes, and maintains a global namespace. With the techniques of system call interception, distributed metadata management, and generic data compression, FanStore provides a POSIX-compliant interface with native hardware throughput in an efficient and scalable manner. Users do not have to make intrusive code changes to use FanStore and take advantage of the optimized I/O. Our experiments with benchmarks and real applications show that FanStore can scale DL training to 512 compute nodes with over 90\% scaling efficiency.

cs.DC↗