arXiv subjects

Shuai Wang

Publications and source records attributed to Shuai Wang.

At least 721 records · Page 40Linked to original sources

arXiv2020

MUTATT: Visual-Textual Mutual Guidance for Referring Expression Comprehension

Referring expression comprehension (REC) aims to localize a text-related region in a given image by a referring expression in natural language. Existing methods focus on how to build convincing visual and language representations independently, which may significantly isolate visual and language information. In this paper, we argue that for REC the referring expression and the target region are semantically correlated and subject, location and relationship consistency exist between vision and language.On top of this, we propose a novel approach called MutAtt to construct mutual guidance between vision and language, which treat vision and language equally thus yield compact information matching. Specifically, for each module of subject, location and relationship, MutAtt builds two kinds of attention-based mutual guidance strategies. One strategy is to generate vision-guided language embedding for the sake of matching relevant visual feature. The other reversely generates language-guided visual feature to match relevant language embedding. This mutual guidance strategy can effectively guarantees the vision-language consistency in three modules. Experiments on three popular REC datasets demonstrate that the proposed approach outperforms the current state-of-the-art methods.

cs.CV↗

arXiv2020

Intelligent Home 3D: Automatic 3D-House Design from Linguistic Descriptions Only

Home design is a complex task that normally requires architects to finish with their professional skills and tools. It will be fascinating that if one can produce a house plan intuitively without knowing much knowledge about home design and experience of using complex designing tools, for example, via natural language. In this paper, we formulate it as a language conditioned visual content generation problem that is further divided into a floor plan generation and an interior texture (such as floor and wall) synthesis task. The only control signal of the generation process is the linguistic expression given by users that describe the house details. To this end, we propose a House Plan Generative Model (HPGM) that first translates the language input to a structural graph representation and then predicts the layout of rooms with a Graph Conditioned Layout Prediction Network (GC LPN) and generates the interior texture with a Language Conditioned Texture GAN (LCT-GAN). With some post-processing, the final product of this task is a 3D house model. To train and evaluate our model, we build the first Text-to-3D House Model dataset.

cs.CV↗

arXiv2020

BUT System for the Second DIHARD Speech Diarization Challenge

This paper describes the winning systems developed by the BUT team for the four tracks of the Second DIHARD Speech Diarization Challenge. For tracks 1 and 2 the systems were mainly based on performing agglomerative hierarchical clustering (AHC) of x-vectors, followed by another x-vector clustering based on Bayes hidden Markov model and variational Bayes inference. We provide a comparison of the improvement given by each step and share the implementation of the core of the system. For tracks 3 and 4 with recordings from the Fifth CHiME Challenge, we explored different approaches for doing multi-channel diarization and our best performance was obtained when applying AHC on the fusion of per channel probabilistic linear discriminant analysis scores.

eess.AS↗

arXiv2020

WatchDog: Real-time Vehicle Tracking on Geo-distributed Edge Nodes

Vehicle tracking, a core application to smart city video analytics, is becoming more widely deployed than ever before thanks to the increasing number of traffic cameras and recent advances of computer vision and machine learning. Due to the constraints of bandwidth, latency, and privacy concerns, tracking tasks are more preferable to run on edge devices sitting close to the cameras. However, edge devices are provisioned with a fixed amount of compute budget, making them incompetent to adapt to time-varying tracking workloads caused by traffic dynamics. In coping with this challenge, we propose WatchDog, a real-time vehicle tracking system fully utilizes edge nodes across the road network. WatchDog leverages computer vision tasks with different resource-accuracy trade-offs, and decompose and schedule tracking tasks judiciously across edge devices based on the current workload to maximize the number of tasks while ensuring a provable response time bound at each edge device. Extensive evaluations have been conducted using real-world city-wide vehicle trajectory datasets, showing a 100% tracking coverage with real-time guarantee.

cs.DC↗

arXiv2020

New Viewpoint and Algorithms for Water-Filling Solutions in Wireless Communications

Water-filling solutions play an important role in the designs for wireless communications, e.g., transmit covariance matrix design. A traditional physical understanding is to use the analogy of pouring water over a pool with fluctuating bottom. Numerous variants of water-filling solutions have been discovered during the evolution of wireless networks. To obtain the solution values, iterative computations are required, even for simple cases with compact mathematical formulations. Thus, algorithm design is a key issue for the practical use of water-filling solutions, which however has been given marginal attention in the literature. Many existing algorithms are designed on a case-by-case basis for the variations of water-filling solutions and/or with complex logics. In this paper, a new viewpoint for water-filling solutions is proposed to understand the problem dynamically by considering changes in the increasing rates on different subchannels. This fresh viewpoint provides useful mechanism and fundamental information in finding the optimization solution values. Based on the new understanding, a novel and comprehensive method for practical water-filling algorithm design is proposed, which can be used for systems with various performance metrics and power constraints, even for systems with imperfect channel state information (CSI).

cs.IT↗

arXiv2020

Hybrid Transceiver Optimization for Multi-Hop Communications

Multi-hop communication with the aid of large-scale antenna arrays will play a vital role in future emergence communication systems. In this paper, we investigate amplify-and-forward based and multiple-input multiple-output assisted multi-hop communication, in which all nodes employ hybrid transceivers. Moreover, channel errors are taken into account in our hybrid transceiver design. Based on the matrix-monotonic optimization framework, the optimal structures of the robust hybrid transceivers are derived. By utilizing these optimal structures, the optimizations of analog transceivers and digital transceivers can be separated without loss of optimality. This fact greatly simplifies the joint optimization of analog and digital transceivers. Since the optimization of analog transceivers under unit-modulus constraints is non-convex, a projection type algorithm is proposed for analog transceiver optimization to overcome this difficulty. Based on the derived analog transceivers, the optimal digital transceivers can then be derived using matrix-monotonic optimization. Numeral results obtained demonstrate the performance advantages of the proposed hybrid transceiver designs over other existing solutions.

cs.IT↗

arXiv2019

Full-color complex-amplitude vectorial holograms based on multi-freedom metasurfaces

Phase, polarization, amplitude and frequency represent the basic dimensions of light, playing crucial roles for both fundamental light-mater interactions and all major optical applications. Metasurface emerges as a compact platform to manipulate these knobs, but previous metasurfaces have limited flexibility to simultaneous control them. Here, we introduce a multi-freedom metasurface that can simultaneously and independently modulate phase, polarization and amplitude in an analytical form, and further realize frequency multiplexing by a k-space engineering technique. The multi-freedom metasurface seamlessly combine geometric Pancharatnam-Berry phase and detour phase, both of which are frequency-independent. As a result, it allows complex-amplitude vectorial hologram at various frequencies based on the same design strategy, without sophisticated nanostructure searching of massive size parameters. Based on this principle, we experimentally demonstrate full-color complex-amplitude vectorial meta-holograms in the visible with a metal-insulator metal architecture, unlocking the long-sought full potential of advanced light field manipulation through ultrathin metasurfaces.

physics.optics↗

arXiv2019

Lifelong and Interactive Learning of Factual Knowledge in Dialogues

Dialogue systems are increasingly using knowledge bases (KBs) storing real-world facts to help generate quality responses. However, as the KBs are inherently incomplete and remain fixed during conversation, it limits dialogue systems' ability to answer questions and to handle questions involving entities or relations that are not in the KB. In this paper, we make an attempt to propose an engine for Continuous and Interactive Learning of Knowledge (CILK) for dialogue systems to give them the ability to continuously and interactively learn and infer new knowledge during conversations. With more knowledge accumulated over time, they will be able to learn better and answer more questions. Our empirical evaluation shows that CILK is promising.

cs.CL↗

arXiv2019

Metamorphic Testing for Object Detection Systems

Recent advances in deep neural networks (DNNs) have led to object detectors that can rapidly process pictures or videos, and recognize the objects that they contain. Despite the promising progress by industrial manufacturers such as Amazon and Google in commercializing deep learning-based object detection as a standard computer vision service, object detection systems - similar to traditional software - may still produce incorrect results. These errors, in turn, can lead to severe negative outcomes for the users of these object detection systems. For instance, an autonomous driving system that fails to detect pedestrians can cause accidents or even fatalities. However, principled, systematic methods for testing object detection systems do not yet exist, despite their importance. To fill this critical gap, we introduce the design and realization of MetaOD, the first metamorphic testing system for object detectors to effectively reveal erroneous detection results by commercial object detectors. To this end, we (1) synthesize natural-looking images by inserting extra object instances into background images, and (2) design metamorphic conditions asserting the equivalence of object detection results between the original and synthetic images after excluding the prediction results on the inserted objects. MetaOD is designed as a streamlined workflow that performs object extraction, selection, and insertion. Evaluated on four commercial object detection services and four pretrained models provided by the TensorFlow API, MetaOD found tens of thousands of detection defects in these object detectors. To further demonstrate the practical usage of MetaOD, we use the synthetic images that cause erroneous detection results to retrain the model. Our results show that the model performance is increased significantly, from an mAP score of 9.3 to an mAP score of 10.5.

cs.CV↗

arXiv2019

Period doubling of multiple dissipative-soliton-resonance pulses in a fiber laser

We report on the experimental observation of period doubling of multiple dissipative-soliton-resonance (DSR) pulses in an all-normal-dispersion fiber laser based on a nonlinear amplifying loop mirror. By increasing the pump power, in addition to the typically linearly pulse broadening under fixed pulse peak power, the jump from a single DSR pulse to multiple DSR pulses was observed. During this process, the period doubling appears, that is, the DSR pulses can exhibit the characteristics of period doubling in a multi-pulse state. The typical DSR performance of linearly pulse duration change versus pump power varying can be maintained when the period doubling of multiple DSR pulses appears.

nlin.PS↗

arXiv2019

Period-doubling bifurcation of dissipative-soliton-resonance pulses in a passively mode-locked fiber laser

We report on the experimental observation of period-doubling bifurcation of dissipative-soliton-resonance (DSR) pulses in a fiber laser passively mode-locked by using the nonlinear optical loop mirror. Increasing the pump power of the fiber laser, we show that temporally a stable, uniform DSR pulse train could be transformed into a period-doubling state, exhibiting two sets of pulse parameters between the adjacent cavity roundtrip. It is found that DSR pulses in the period-doubling state could maintain the typical feature of DSR pulse: fixed pulse peak power and linear variation in pulse width with respect to the pump power change. The mechanism for achieving period-doubling of DSR pulses is discussed.

physics.optics↗

arXiv2019

BUT System Description for DIHARD Speech Diarization Challenge 2019

This paper describes the systems developed by the BUT team for the four tracks of the second DIHARD speech diarization challenge. For tracks 1 and 2 the systems were based on performing agglomerative hierarchical clustering (AHC) over x-vectors, followed by the Bayesian Hidden Markov Model (HMM) with eigenvoice priors applied at x-vector level followed by the same approach applied at frame level. For tracks 3 and 4, the systems were based on performing AHC using x-vectors extracted on all channels.

eess.AS↗

arXiv2019

BUT System Description to VoxCeleb Speaker Recognition Challenge 2019

In this report, we describe the submission of Brno University of Technology (BUT) team to the VoxCeleb Speaker Recognition Challenge (VoxSRC) 2019. We also provide a brief analysis of different systems on VoxCeleb-1 test sets. Submitted systems for both Fixed and Open conditions are a fusion of 4 Convolutional Neural Network (CNN) topologies. The first and second networks have ResNet34 topology and use two-dimensional CNNs. The last two networks are one-dimensional CNN and are based on the x-vector extraction topology. Some of the networks are fine-tuned using additive margin angular softmax. Kaldi FBanks and Kaldi PLPs were used as features. The difference between Fixed and Open systems lies in the used training data and fusion strategy. The best systems for Fixed and Open conditions achieved 1.42% and 1.26% ERR on the challenge evaluation set respectively.

eess.AS↗

arXiv2019

Studies on ion back-flow of Time Projection Chamber based on GEM and anode wire grid

Gated wires are widely used in Time Projection Chamber (TPC) to avoid ion back-flow (IBF) in the drift volume. The anode wires can provide stable gain at high voltage with a long lifetime. However, switching on and off the gated grid (GG) leads to a dead time and also limit the readout efficiency of the TPC. Gas Electron Multiplier (GEM) foil provides a possibility of continuous readout for TPC, which can suppress IBF efficiently while keeping stable gain. A prototype chamber including two layers of GEM foils and anode wires has been built to combine both advantages from GEM and anode wire. Using Garfield++ and the finite element analysis (FEA) method, simulations of the transmission processes of electrons and ions are performed and results on absorption ratio of ions, gain and IBF ratio are obtained. The optimized parameters from simulation are then applied to the prototype chamber to test the IBF and other performances. Both GEM foils are run at low voltage (255V), while most of the gain is provided by the anode wire. The measurement shows that the IBF ratio can be suppressed to ~0.58% with double-layer GEM foils (staggered) at an effective gain about 2500 with an energy resolution about 10%.

physics.ins-det↗

arXiv2019

Suggestion of the DLV dimensionless number system to represent the scaled behavior of structures under impact loads

A group of dimensionless numbers, termed DLV (Density-Length-Velocity) system, is put forward to represent the scaled behavior of structures under impact loads. It is obtained by means of the Buckingham Pi theorem with an alternative basis. The distinct features of this group of dimensionless numbers are that it relates physical quantities of the impacted structure with essential basis of the Density, the Length and the Velocity, and thus it can represent the scaled influence of material property, geometry characteristic and velocity on the behavior of structures. The newly 15 proposed dimensionless numbers reflect three advantages. (1) The intuitively clear physical significance of these dimensionless numbers, such as the ratios of force intensity, force, moment of inertia to the corresponding dynamic quantities, the Johnson's damage number Dn and Zhao's response number Rn etc. are naturally included. (2) The property of direct matching the dimensionless expression of response equations of dynamic problems with these dimensionless numbers through simple equation analysis; (3) The ability of addressing non-scaling problems for different materials and strain-rate-sensitive as well as the VSG (initial impact Velocity-dynamic flow Stress-impact mass G) system. Four classical impact models are used to verify the direct matching property and the non-scaling addressing ability of the DLV system by equation analysis. The results show that the proposed dimensionless number system is simple, clear and efficient, and we suggest using it to represent the scaled behavior of structures under impact loads.

physics.class-ph↗

arXiv2019

Margin Matters: Towards More Discriminative Deep Neural Network Embeddings for Speaker Recognition

Recently, speaker embeddings extracted from a speaker discriminative deep neural network (DNN) yield better performance than the conventional methods such as i-vector. In most cases, the DNN speaker classifier is trained using cross entropy loss with softmax. However, this kind of loss function does not explicitly encourage inter-class separability and intra-class compactness. As a result, the embeddings are not optimal for speaker recognition tasks. In this paper, to address this issue, three different margin based losses which not only separate classes but also demand a fixed margin between classes are introduced to deep speaker embedding learning. It could be demonstrated that the margin is the key to obtain more discriminative speaker embeddings. Experiments are conducted on two public text independent tasks: VoxCeleb1 and Speaker in The Wild (SITW). The proposed approach can achieve the state-of-the-art performance, with 25% ~ 30% equal error rate (EER) reduction on both tasks when compared to strong baselines using cross entropy loss with softmax, obtaining 2.238% EER on VoxCeleb1 test set and 2.761% EER on SITW core-core test set, respectively.

eess.AS↗

arXiv2019

Forward and Backward Knowledge Transfer for Sentiment Classification

This paper studies the problem of learning a sequence of sentiment classification tasks. The learned knowledge from each task is retained and used to help future or subsequent task learning. This learning paradigm is called Lifelong Learning (LL). However, existing LL methods either only transfer knowledge forward to help future learning and do not go back to improve the model of a previous task or require the training data of the previous task to retrain its model to exploit backward/reverse knowledge transfer. This paper studies reverse knowledge transfer of LL in the context of naive Bayesian (NB) classification. It aims to improve the model of a previous task by leveraging future knowledge without retraining using its training data. This is done by exploiting a key characteristic of the generative model of NB. That is, it is possible to improve the NB classifier for a task by improving its model parameters directly by using the retained knowledge from other tasks. Experimental results show that the proposed method markedly outperforms existing LL baselines.

cs.AI↗

arXiv2019

Identifying Cache-Based Side Channels through Secret-Augmented Abstract Interpretation

Cache-based side channels enable a dedicated attacker to reveal program secrets by measuring the cache access patterns. Practical attacks have been shown against real-world crypto algorithm implementations such as RSA, AES, and ElGamal. By far, identifying information leaks due to cache-based side channels, either in a static or dynamic manner, remains a challenge: the existing approaches fail to offer high precision, full coverage, and good scalability simultaneously, thus impeding their practical use in real-world scenarios. In this paper, we propose a novel static analysis method on binaries to detect cache-based side channels. We use abstract interpretation to reason on program states with respect to abstract values at each program point. To make such abstract interpretation scalable to real-world cryptosystems while offering high precision and full coverage, we propose a novel abstract domain called the Secret-Augmented Symbolic domain (SAS). SAS tracks program secrets and dependencies on them for precision, while it tracks only coarse-grained public information for scalability. We have implemented the proposed technique into a practical tool named CacheS and evaluated it on the implementations of widely-used cryptographic algorithms in real-world crypto libraries, including Libgcrypt, OpenSSL, and mbedTLS. CacheS successfully confirmed a total of 154 information leaks reported by previous research and 54 leaks that were previously unknown. We have reported our findings to the developers. And they confirmed that many of those unknown information leaks do lead to potential side channels.

cs.CR↗