Searcharxiv⌕ Search

arXiv subjects

Hao Zhang

Publications and source records attributed to Hao Zhang.

At least 1,063 records · Page 59Linked to original sources
arXiv2021

MetaSCI: Scalable and Adaptive Reconstruction for Video Compressive Sensing

To capture high-speed videos using a two-dimensional detector, video snapshot compressive imaging (SCI) is a promising system, where the video frames are coded by different masks and then compressed to a snapshot measurement. Following this, efficient algorithms are desired to reconstruct the high-speed frames, where the state-of-the-art results are achieved by deep learning networks. However, these networks are usually trained for specific small-scale masks and often have high demands of training time and GPU memory, which are hence {\bf \em not flexible} to $i$) a new mask with the same size and $ii$) a larger-scale mask. We address these challenges by developing a Meta Modulated Convolutional Network for SCI reconstruction, dubbed MetaSCI. MetaSCI is composed of a shared backbone for different masks, and light-weight meta-modulation parameters to evolve to different modulation parameters for each mask, thus having the properties of {\bf \em fast adaptation} to new masks (or systems) and ready to {\bf \em scale to large data}. Extensive simulation and real data results demonstrate the superior performance of our proposed approach. Our code is available at {\small\url{https://github.com/xyvirtualgroup/MetaSCI-CVPR2021}}.

eess.IV↗
arXiv2021

Natural Language Video Localization: A Revisit in Span-based Question Answering Framework

Natural Language Video Localization (NLVL) aims to locate a target moment from an untrimmed video that semantically corresponds to a text query. Existing approaches mainly solve the NLVL problem from the perspective of computer vision by formulating it as ranking, anchor, or regression tasks. These methods suffer from large performance degradation when localizing on long videos. In this work, we address the NLVL from a new perspective, i.e., span-based question answering (QA), by treating the input video as a text passage. We propose a video span localizing network (VSLNet), on top of the standard span-based QA framework (named VSLBase), to address NLVL. VSLNet tackles the differences between NLVL and span-based QA through a simple yet effective query-guided highlighting (QGH) strategy. QGH guides VSLNet to search for the matching video span within a highlighted region. To address the performance degradation on long videos, we further extend VSLNet to VSLNet-L by applying a multi-scale split-and-concatenation strategy. VSLNet-L first splits the untrimmed video into short clip segments; then, it predicts which clip segment contains the target moment and suppresses the importance of other segments. Finally, the clip segments are concatenated, with different confidences, to locate the target moment accurately. Extensive experiments on three benchmark datasets show that the proposed VSLNet and VSLNet-L outperform the state-of-the-art methods; VSLNet-L addresses the issue of performance degradation on long videos. Our study suggests that the span-based QA framework is an effective strategy to solve the NLVL problem.

cs.CL↗
arXiv2021

Dictionary Learning with Low-rank Coding Coefficients for Tensor Completion

In this paper, we propose a novel tensor learning and coding model for third-order data completion. Our model is to learn a data-adaptive dictionary from the given observations, and determine the coding coefficients of third-order tensor tubes. In the completion process, we minimize the low-rankness of each tensor slice containing the coding coefficients. By comparison with the traditional pre-defined transform basis, the advantages of the proposed model are that (i) the dictionary can be learned based on the given data observations so that the basis can be more adaptively and accurately constructed, and (ii) the low-rankness of the coding coefficients can allow the linear combination of dictionary features more effectively. Also we develop a multi-block proximal alternating minimization algorithm for solving such tensor learning and coding model, and show that the sequence generated by the algorithm can globally converge to a critical point. Extensive experimental results for real data sets such as videos, hyperspectral images, and traffic data are reported to demonstrate these advantages and show the performance of the proposed tensor learning and coding method is significantly better than the other tensor completion methods in terms of several evaluation metrics.

cs.CV↗
arXiv2021

Fast Distributed Complex Join Processing

In this work, we study the problem of co-optimize communication, pre-computing, and computation cost in one-round multi-way join evaluation. We propose a multi-way join approach ADJ (Adaptive Distributed Join) for complex join which finds one optimal query plan to process by exploring cost-effective partial results in terms of the trade-off between pre-computing, communication, and computation.We analyze the input relations for a given join query and find one optimal over a set of query plans in some specific form, with high-quality cost estimation by sampling. Our extensive experiments confirm that ADJ outperforms the existing multi-way join methods by up to orders of magnitude.

cs.DB↗
arXiv2021

Interpreting Multivariate Shapley Interactions in DNNs

This paper aims to explain deep neural networks (DNNs) from the perspective of multivariate interactions. In this paper, we define and quantify the significance of interactions among multiple input variables of the DNN. Input variables with strong interactions usually form a coalition and reflect prototype features, which are memorized and used by the DNN for inference. We define the significance of interactions based on the Shapley value, which is designed to assign the attribution value of each input variable to the inference. We have conducted experiments with various DNNs. Experimental results have demonstrated the effectiveness of the proposed method.

cs.LG↗
arXiv2021

Fast Dynamics in a Model Metallic Glass-forming Material

We investigate the fast $β$- and Johari-Goldstein (JG) $β$-relaxation processes, along with the elastic scattering response of glass-forming (GF) liquids and the Boson peak, in a simulated Al-Sm GF material exhibiting a fragile-strong (FS) transition. These dynamical processes are universal in 'ordinary' GF fluids and collectively describe their 'fast dynamics', and we find these relaxation processes also arise in a GF liquid exhibiting a fragile-strong transition. String-like particle motion, having both an irreversible and reversible nature ('stringlets') component, occurs in the fast-dynamics regime, corresponding to a ps timescale. String-like collective motion associated with localized unstable modes facilitate irreversible and intermittent particle 'jumping' events at long times associated with the JG $β$-relaxation process, while stringlets associated with localized stable modes and corresponding perfectly reversible atomic motion give rise to the Boson peak. To further clarify the origin of the Boson peak, we calculate the density of states for both the stringlet particles and the 'normal' particles and find that the stringlet particles give rise to a Boson peak while the normal atoms do not. The growth of stringlets upon heating ultimately also leads to the 'softening' of these excitations, and the Boson peak frequency and shear modulus drop in concert with this softening. The growth of string-like collective motion upon heating in the fast-dynamics regime is further shown to be responsible for the growth in the intensity of the fast relaxation process. Relaxation in cooled liquids clearly involves a hierarchy of relaxation processes acting on rather different time and spatial scales.

cond-mat.mtrl-sci↗
arXiv2021

Large zero-bias peaks in InSb-Al hybrid semiconductor-superconductor nanowire devices

We report electron transport studies on InSb-Al hybrid semiconductor-superconductor nanowire devices. Tunnelling spectroscopy is used to measure the evolution of subgap states while varying magnetic field and voltages applied to various nearby gates. At magnetic fields between 0.7 and 0.9 T, the differential conductance contains large zero bias peaks (ZBPs) whose height reaches values on the order 2e2/h. We investigate these ZBPs for large ranges of gate voltages in different devices. We discuss possible interpretations in terms of disorder-induced subgap states, Andreev bound states and Majorana zero modes.

cond-mat.mes-hall↗
arXiv2021

Data Association Between Perception and V2V Communication Sensors

The connectivity between vehicles, infrastructure, and other traffic participants brings a new dimension to automotive safety applications. Soon all the newly produced cars will have Vehicle to Everything (V2X) communication modems alongside the existing Advanced Driver Assistant Systems (ADAS). It is essential to identify the different sensor measurements for the same targets (Data Association) to use connectivity reliably as a safety feature alongside the standard ADAS functionality. Considering the camera is the most common sensor available for ADAS systems, in this paper, we present an experimental implementation of a Mahalanobis distance-based data association algorithm between the camera and the Vehicle to Vehicle (V2V) communication sensors. The implemented algorithm has low computational complexity and the capability of running in real-time. One can use the presented algorithm for sensor fusion algorithms or higher-level decision-making applications in ADAS modules.

cs.RO↗
arXiv2021

Reinforcement Learning for Flexibility Design Problems

Flexibility design problems are a class of problems that appear in strategic decision-making across industries, where the objective is to design a ($e.g.$, manufacturing) network that affords flexibility and adaptivity. The underlying combinatorial nature and stochastic objectives make flexibility design problems challenging for standard optimization methods. In this paper, we develop a reinforcement learning (RL) framework for flexibility design problems. Specifically, we carefully design mechanisms with noisy exploration and variance reduction to ensure empirical success and show the unique advantage of RL in terms of fast-adaptation. Empirical results show that the RL-based method consistently finds better solutions compared to classical heuristics.

cs.LG↗
arXiv2021

Building Interpretable Interaction Trees for Deep NLP Models

This paper proposes a method to disentangle and quantify interactions among words that are encoded inside a DNN for natural language processing. We construct a tree to encode salient interactions extracted by the DNN. Six metrics are proposed to analyze properties of interactions between constituents in a sentence. The interaction is defined based on Shapley values of words, which are considered as an unbiased estimation of word contributions to the network prediction. Our method is used to quantify word interactions encoded inside the BERT, ELMo, LSTM, CNN, and Transformer networks. Experimental results have provided a new perspective to understand these DNNs, and have demonstrated the effectiveness of our method.

cs.CL↗
arXiv2021

Spectral Deconvolution Analysis on Olivine-Orthopyroxene Mixtures with Simulated Space Weathering Modifications

Olivine and pyroxene are important mineral end-members for studying the sur-face material compositions of mafic bodies. The profiles of visible and near-infraredspectra of olivine-orthopyroxene mixtures systematically varied with their compositionratios. In our experiments, we combine the RELAB spectral database with a new spec-tral data obtained from some assembled olivine-orthopyroxene mixtures. We found thatthe commonly-used band area ratio (BAR, Cloutis et al. 1986) does not work well onour newly obtained spectral data. To investigate this issue, an empirical procedure basedon fitted results by modified Gaussian model is proposed to analyze the spectral curves.Following the new empirical procedure, the end-member abundances can be estimatedwith a 15% accuracy with some prior mineral absorption features. In addition, the mix-ture samples configured in our experiments are also irradiated by pulsed lasers to simulateand investigate the space weathering effects. Spectral deconvolution results confirm thatlow-content olivine on celestial bodies are difficult to measure and estimate. Therefore,the olivine abundance of space weathered materials may be underestimated from remotesensing data. This study may be used to quantify the spectral relationship of olivine-orthopyroxene mixtures and further reveal their correlation between the spectra of ordi-nary chondrites and silicate asteroids.

astro-ph.EP↗
arXiv2021

Quantum annealing simulation of out-of-equilibrium magnetization in a spin-chain compound

Geometrically frustrated spin-chain compounds such as Ca3Co2O6 exhibit extremely slow relaxation under a changing magnetic field. Consequently, both low-temperature laboratory experiments and Monte Carlo simulations have shown peculiar out-of-equilibrium magnetization curves, which arise from trapping in metastable configurations. In this work we simulate this phenomenon in a superconducting quantum annealing processor, allowing us to probe the impact of quantum fluctuations on both equilibrium and dynamics of the system. Increasing the quantum fluctuations with a transverse field reduces the impact of metastable traps in out-of-equilibrium samples, and aids the development of three-sublattice ferrimagnetic (up-up-down) long-range order. At equilibrium we identify a finite-temperature shoulder in the 1/3-to-saturated phase transition, promoted by quantum fluctuations but with entropic origin. This work demonstrates the viability of dynamical as well as equilibrium studies of frustrated magnetism using large-scale programmable quantum systems, and is therefore an important step toward programmable simulation of dynamics in materials using quantum hardware.

quant-ph↗
arXiv2021

Robot Adaptation for Generating Consistent Navigational Behaviors over Unstructured Off-Road Terrain

Terrain adaptation is an essential capability for a ground robot to effectively traverse unstructured off-road terrain in real-world field environments such as forests. However, the expected robot behaviors generated by terrain adaptation methods cannot always be executed accurately due to setbacks such as wheel slip and reduced tire pressure. To address this problem, we propose a novel approach for consistent behavior generation that enables the ground robot's actual behaviors to more accurately match expected behaviors while adapting to a variety of unstructured off-road terrain. Our approach learns offset behaviors that are used to compensate for the inconsistency between the actual and expected behaviors without requiring the explicit modeling of various setbacks. Our approach is also able to estimate the importance of the multi-modal features to improve terrain representations for better adaptation. In addition, we develop an algorithmic solver for our formulated regularized optimization problem, which is guaranteed to converge to the global optimal solution. To evaluate the method, we perform extensive experiments using various unstructured off-road terrain in real-world field environments. Experimental results have validated that our approach enables robots to traverse complex unstructured off-road terrain with more navigational behavior consistency, and it outperforms previous methods, particularly so on challenging terrain.

cs.RO↗
arXiv2020

D$^2$IM-Net: Learning Detail Disentangled Implicit Fields from Single Images

We present the first single-view 3D reconstruction network aimed at recovering geometric details from an input image which encompass both topological shape structures and surface features. Our key idea is to train the network to learn a detail disentangled reconstruction consisting of two functions, one implicit field representing the coarse 3D shape and the other capturing the details. Given an input image, our network, coined D$^2$IM-Net, encodes it into global and local features which are respectively fed into two decoders. The base decoder uses the global features to reconstruct a coarse implicit field, while the detail decoder reconstructs, from the local features, two displacement maps, defined over the front and back sides of the captured object. The final 3D reconstruction is a fusion between the base shape and the displacement maps, with three losses enforcing the recovery of coarse shape, overall structure, and surface details via a novel Laplacian term.

cs.CV↗
arXiv2020

Unsupervised Image Segmentation using Mutual Mean-Teaching

Unsupervised image segmentation aims at assigning the pixels with similar feature into a same cluster without annotation, which is an important task in computer vision. Due to lack of prior knowledge, most of existing model usually need to be trained several times to obtain suitable results. To address this problem, we propose an unsupervised image segmentation model based on the Mutual Mean-Teaching (MMT) framework to produce more stable results. In addition, since the labels of pixels from two model are not matched, a label alignment algorithm based on the Hungarian algorithm is proposed to match the cluster labels. Experimental results demonstrate that the proposed model is able to segment various types of images and achieves better performance than the existing methods.

cs.CV↗
arXiv2020

Mastering Complex Control in MOBA Games with Deep Reinforcement Learning

We study the reinforcement learning problem of complex action control in the Multi-player Online Battle Arena (MOBA) 1v1 games. This problem involves far more complicated state and action spaces than those of traditional 1v1 games, such as Go and Atari series, which makes it very difficult to search any policies with human-level performance. In this paper, we present a deep reinforcement learning framework to tackle this problem from the perspectives of both system and algorithm. Our system is of low coupling and high scalability, which enables efficient explorations at large scale. Our algorithm includes several novel strategies, including control dependency decoupling, action mask, target attention, and dual-clip PPO, with which our proposed actor-critic network can be effectively trained in our system. Tested on the MOBA game Honor of Kings, our AI agent, called Tencent Solo, can defeat top professional human players in full 1v1 games.

cs.AI↗
arXiv2020

Deep N-ary Error Correcting Output Codes

Ensemble learning consistently improves the performance of multi-class classification through aggregating a series of base classifiers. To this end, data-independent ensemble methods like Error Correcting Output Codes (ECOC) attract increasing attention due to its easiness of implementation and parallelization. Specifically, traditional ECOCs and its general extension N-ary ECOC decompose the original multi-class classification problem into a series of independent simpler classification subproblems. Unfortunately, integrating ECOCs, especially N-ary ECOC with deep neural networks, termed as deep N-ary ECOC, is not straightforward and yet fully exploited in the literature, due to the high expense of training base learners. To facilitate the training of N-ary ECOC with deep learning base learners, we further propose three different variants of parameter sharing architectures for deep N-ary ECOC. To verify the generalization ability of deep N-ary ECOC, we conduct experiments by varying the backbone with different deep neural network architectures for both image and text classification tasks. Furthermore, extensive ablation studies on deep N-ary ECOC show its superior performance over other deep data-independent ensemble methods.

cs.CV↗