arXiv subjects

Dong Wang

Publications and source records attributed to Dong Wang.

At least 649 records · Page 36Linked to original sources

arXiv2019

VAE-based Domain Adaptation for Speaker Verification

Deep speaker embedding has achieved satisfactory performance in speaker verification. By enforcing the neural model to discriminate the speakers in the training set, deep speaker embedding (called `x-vectors`) can be derived from the hidden layers. Despite its good performance, the present embedding model is highly domain sensitive, which means that it often works well in domains whose acoustic condition matches that of the training data (in-domain), but degrades in mismatched domains (out-of-domain). In this paper, we present a domain adaptation approach based on Variational Auto-Encoder (VAE). This model transforms x-vectors to a regularized latent space; within this latent space, a small amount of data from the target domain is sufficient to accomplish the adaptation. Our experiments demonstrated that by this VAE-adaptation approach, speaker embeddings can be easily transformed to the target domain, leading to noticeable performance improvement.

eess.AS↗

arXiv2019

Quantum-memory-assisted entropic uncertainty relations

Uncertainty relations take a crucial and fundamental part in the frame of quantum theory, and are bringing on many marvelous applications in the emerging field of quantum information sciences. Especially, as entropy is imposed into the uncertainty principle, entropy-based uncertainty relations lead to a number of applications including quantum key distribution, entanglement witness, quantum steering, quantum metrology, and quantum teleportation. Herein, the history of the development of the uncertainty relations is discussed, especially focusing on the recent progress with regard to quantum-memory-assisted entropic uncertainty relations and dynamical characteristics of the measured uncertainty in some explicit physical systems. The aims are to help deepen the understanding of entropic uncertainty relations and prompt further explorations for versatile applications of the relations on achieving practical quantum tasks.

quant-ph↗

arXiv2019

Noninvasive super-resolution imaging through scattering media

Super-resolution imaging with advanced optical systems has been revolutionizing technical analysis in various fields from biological to physical sciences. However, many objects are hidden by strongly scattering media such as rough wall corners or biological tissues that scramble light paths, create speckle patterns and hinder object's visualization, let alone super-resolution imaging. Here, we realize a method to do non-invasive super-resolution imaging through scattering media based on stochastic optical scattering localization imaging (SOSLI) technique. Simply by capturing multiple speckle patterns of photo-switchable emitters in our demonstration, the stochastic approach utilizes the speckle correlation properties of scattering media to retrieve an image with more than five-fold resolution enhancement compared to the diffraction limit, while posing no fundamental limit in achieving higher spatial resolution. More importantly, we demonstrate our SOSLI to do non-invasive super-resolution imaging through not only optical diffusers, i.e. static scattering media, but also biological tissues, i.e. dynamic scattering media with decorrelation of up to 80%. Our approach paves the way to non-invasively visualize various samples behind scattering media at unprecedented levels of detail.

physics.optics↗

arXiv2019

signADAM: Learning Confidences for Deep Neural Networks

In this paper, we propose a new first-order gradient-based algorithm to train deep neural networks. We first introduce the sign operation of stochastic gradients (as in sign-based methods, e.g., SIGN-SGD) into ADAM, which is called as signADAM. Moreover, in order to make the rate of fitting each feature closer, we define a confidence function to distinguish different components of gradients and apply it to our algorithm. It can generate more sparse gradients than existing algorithms do. We call this new algorithm signADAM++. In particular, both our algorithms are easy to implement and can speed up training of various deep neural networks. The motivation of signADAM++ is preferably learning features from the most different samples by updating large and useful gradients regardless of useless information in stochastic gradients. We also establish theoretical convergence guarantees for our algorithms. Empirical results on various datasets and models show that our algorithms yield much better performance than many state-of-the-art algorithms including SIGN-SGD, SIGNUM and ADAM. We also analyze the performance from multiple perspectives including the loss landscape and develop an adaptive method to further improve generalization. The source code is available at https://github.com/DongWanginxdu/signADAM-Learn-by-Confidence.

cs.CV↗

arXiv2019

Towards Reliable Online Clickbait Video Detection: A Content-Agnostic Approach

Online video sharing platforms (e.g., YouTube, Vimeo) have become an increasingly popular paradigm for people to consume video contents. Clickbait video, whose content clearly deviates from its title/thumbnail, has emerged as a critical problem on online video sharing platforms. Current clickbait detection solutions that mainly focus on analyzing the text of the title, the image of the thumbnail, or the content of the video are shown to be suboptimal in detecting the online clickbait videos. In this paper, we develop a novel content-agnostic scheme, Online Video Clickbait Protector (OVCP), to effectively detect clickbait videos by exploring the comments from the audience who watched the video. Different from existing solutions, OVCP does not directly analyze the content of the video and its pre-click information (e.g., title and thumbnail). Therefore, it is robust against sophisticated content creators who often generate clickbait videos that can bypass the current clickbait detectors. We evaluate OVCP with a real-world dataset collected from YouTube. Experimental results demonstrate that OVCP is effective in identifying clickbait videos and significantly outperforms both state-of-the-art baseline models and human annotators.

cs.SI↗

arXiv2019

C^3 Framework: An Open-source PyTorch Code for Crowd Counting

This technical report attempts to provide efficient and solid kits addressed on the field of crowd counting, which is denoted as Crowd Counting Code Framework (C$^3$F). The contributions of C$^3$F are in three folds: 1) Some solid baseline networks are presented, which have achieved the state-of-the-arts. 2) Some flexible parameter setting strategies are provided to further promote the performance. 3) A powerful log system is developed to record the experiment process, which can enhance the reproducibility of each experiment. Our code is made publicly available at \url{https://github.com/gjy3035/C-3-Framework}. Furthermore, we also post a Chinese blog\footnote{\url{https://zhuanlan.zhihu.com/p/65650998}} to describe the details and insights of crowd counting.

cs.CV↗

arXiv2019

LMVP: Video Predictor with Leaked Motion Information

We propose a Leaked Motion Video Predictor (LMVP) to predict future frames by capturing the spatial and temporal dependencies from given inputs. The motion is modeled by a newly proposed component, motion guider, which plays the role of both learner and teacher. Specifically, it {\em learns} the temporal features from real data and {\em guides} the generator to predict future frames. The spatial consistency in video is modeled by an adaptive filtering network. To further ensure the spatio-temporal consistency of the prediction, a discriminator is also adopted to distinguish the real and generated frames. Further, the discriminator leaks information to the motion guider and the generator to help the learning of motion. The proposed LMVP can effectively learn the static and temporal features in videos without the need for human labeling. Experiments on synthetic and real data demonstrate that LMVP can yield state-of-the-art results.

cs.CV↗

arXiv2019

A One-step Pruning-recovery Framework for Acceleration of Convolutional Neural Networks

Acceleration of convolutional neural network has received increasing attention during the past several years. Among various acceleration techniques, filter pruning has its inherent merit by effectively reducing the number of convolution filters. However, most filter pruning methods resort to tedious and time-consuming layer-by-layer pruning-recovery strategy to avoid a significant drop of accuracy. In this paper, we present an efficient filter pruning framework to solve this problem. Our method accelerates the network in one-step pruning-recovery manner with a novel optimization objective function, which achieves higher accuracy with much less cost compared with existing pruning methods. Furthermore, our method allows network compression with global filter pruning. Given a global pruning rate, it can adaptively determine the pruning rate for each single convolutional layer, while these rates are often set as hyper-parameters in previous approaches. Evaluated on VGG-16 and ResNet-50 using ImageNet, our approach outperforms several state-of-the-art methods with less accuracy drop under the same and even much fewer floating-point operations (FLOPs).

cs.CV↗

arXiv2019

Interface dynamics for an Allen-Cahn-type equation governing a matrix-valued field

We consider the initial value problem for the generalized Allen-Cahn equation, \[\partial_t Φ= ΔΦ-\varepsilon^{-2} Φ(Φ^t Φ- I), \qquad x \in Ω, \ t\geq 0,\] where $Φ$ is an $n\times n$ real matrix-valued field, $Ω$ is a two-dimensional square with periodic boundary conditions, and $\varepsilon > 0$. This equation is the gradient flow for the energy, $E(Φ) := \int \frac{1}{2} \|\nabla Φ\|^2_F+\frac{1}{4 \varepsilon^2} \| Φ^t Φ- I \|^2_F$, where $\| \cdot \|_F$ denotes the Frobenius norm. The primary contribution of this paper is to use asymptotic methods to describe the solution of this initial value problem. If the initial condition has single-signed determinant, at each point of the domain, at a fast $O(\varepsilon^{-2} t)$ time scale, the solution evolves towards the closest orthogonal matrix. Then, at the $O(t)$ time scale, the solution evolves according to the $O_n$ diffusion equation. Stationary solutions to the $O_n$ diffusion equation are analyzed for $n=2$. If the initial condition has regions where the determinant is positive and negative, a free interface develops. Away from the interface, in each region, the matrix-valued field behaves as in the single-signed determinant case. At the $O(t)$ time scale, the interface evolves in the normal direction by curvature. At a slow $O(\varepsilon t)$ time scale, the interface is driven by curvature and the surface diffusion of the matrix-valued field. For $n=2$, the interface is driven by curvature and the jump in the squared tangental derivative of the phase across the interface. In particular, we emphasize that the interface when $n\geq 2$ is driven by surface diffusion, while for $n=1$, the original Allen--Cahn equation, the interface is only driven by mean curvature. A variety of numerical experiments are performed to verify, support, and illustrate our analytical results.

math.AP↗

arXiv2019

A Preliminary Study on Data Augmentation of Deep Learning for Image Classification

Deep learning models have a large number of freeparameters that need to be calculated by effective trainingof the models on a great deal of training data to improvetheir generalization performance. However, data obtaining andlabeling is expensive in practice. Data augmentation is one of themethods to alleviate this problem. In this paper, we conduct apreliminary study on how three variables (augmentation method,augmentation rate and size of basic dataset per label) can affectthe accuracy of deep learning for image classification. The studyprovides some guidelines: (1) it is better to use transformationsthat alter the geometry of the images rather than those justlighting and color. (2) 2-3 times augmentation rate is good enoughfor training. (3) the smaller amount of data, the more obviouscontributions could have.

cs.CV↗

arXiv2019

Exploiting Persona Information for Diverse Generation of Conversational Responses

In human conversations, due to their personalities in mind, people can easily carry out and maintain the conversations. Giving conversational context with persona information to a chatbot, how to exploit the information to generate diverse and sustainable conversations is still a non-trivial task. Previous work on persona-based conversational models successfully make use of predefined persona information and have shown great promise in delivering more realistic responses. And they all learn with the assumption that given a source input, there is only one target response. However, in human conversations, there are massive appropriate responses to a given input message. In this paper, we propose a memory-augmented architecture to exploit persona information from context and incorporate a conditional variational autoencoder model together to generate diverse and sustainable conversations. We evaluate the proposed model on a benchmark persona-chat dataset. Both automatic and human evaluations show that our model can deliver more diverse and more engaging persona-based responses than baseline approaches.

cs.AI↗

arXiv2019

When Social Sensing Meets Edge Computing: Vision and Challenges

This paper overviews the state of the art, research challenges, and future opportunities in an emerging research direction: Social Sensing based Edge Computing (SSEC). Social sensing has emerged as a new sensing application paradigm where measurements about the physical world are collected from humans or from devices on their behalf. The advent of edge computing pushes the frontier of computation, service, and data along the cloud-to-things continuum. The merging of these two technical trends generates a set of new research challenges that need to be addressed. In this paper, we first define the new SSEC paradigm that is motivated by a few underlying technology trends. We then present a few representative real-world case studies of SSEC applications and several key research challenges that exist in those applications. Finally, we envision a few exciting research directions in future SSEC. We hope this paper will stimulate discussions of this emerging research direction in the community.

cs.HC↗

arXiv2019

Experimental investigation of nonlocal advantage of quantum coherence

Nonlocal advantage of quantum coherence (NAQC) based on coherence complementarity relations is generally viewed as a stronger nonclassical correlation than Bell nonlocality. An arbitrary two-qubit state with NAQC must be an entangled state, which demonstrates that the criterion of NAQC can also be regarded as an entanglement witness. In this paper, we experimentally investigate the NAQC for Bell-diagonal states with high fidelity in an optics-based platform. We perform local measurements on a subsystem in three mutually unbiased bases and reconstruct the density matrices of the measured states by quantum state tomography process. By analyzing characteristic of the $l_1$ norm, relative entropy and skew information of coherence with parameters of quantum states, NAQC for the quantum states is accurately captured, and it shows that our experimental results are well compatible with the theoretical predictions. It is worth mentioning that quantum states with NAQC would have higher entanglement, and thus NAQC could be expected to be a kind of useful physical resource for quantum information processing.

quant-ph↗

arXiv2019

Why torus-unstable solar filaments experience failed eruption?

To investigate the factors that control the success and/or failure of solar eruptions, we study the magnetic field and 3-Dimensional (3D) configuration of 16 filament eruptions during 2010 July - 2013 February. All these events, i.e., erupted but failed to be ejected to become a coronal mass ejection (CME), are failed eruptions with the filament maximum height exceeding $100 Mm$. The magnetic field of filament source regions is approximated by a potential field extrapolation method. The filament 3D configuration is reconstructed from three vantage points by the observations of STEREO Ahead/Behind and SDO spacecraft. We calculate the decay index at the apex of these failed filaments and find that in 7 cases, their apex decay indexes exceed the theoretical threshold ($n_{crit} = 1.5$) of the torus instability. We further determine the orientation change or rotation angle of each filament top during the eruption. Finally, the distribution of these events in the parameter space of rotation angle versus decay index is established. Four distinct regimes in the parameter space are empirically identified. We find that, all the torus-unstable cases (decay index $n > 1.5$), have a large rotation angles ranging from $50^\circ - 130^\circ$. The possible mechanisms leading to the rotation and failed eruption are discussed. These results imply that, besides the torus instability, the rotation motion during the eruption may also play a significant role in solar eruptions.

astro-ph.SR↗

arXiv2019

Memory-Augmented Temporal Dynamic Learning for Action Recognition

Human actions captured in video sequences contain two crucial factors for action recognition, i.e., visual appearance and motion dynamics. To model these two aspects, Convolutional and Recurrent Neural Networks (CNNs and RNNs) are adopted in most existing successful methods for recognizing actions. However, CNN based methods are limited in modeling long-term motion dynamics. RNNs are able to learn temporal motion dynamics but lack effective ways to tackle unsteady dynamics in long-duration motion. In this work, we propose a memory-augmented temporal dynamic learning network, which learns to write the most evident information into an external memory module and ignore irrelevant ones. In particular, we present a differential memory controller to make a discrete decision on whether the external memory module should be updated with current feature. The discrete memory controller takes in the memory history, context embedding and current feature as inputs and controls information flow into the external memory module. Additionally, we train this discrete memory controller using straight-through estimator. We evaluate this end-to-end system on benchmark datasets (UCF101 and HMDB51) of human action recognition. The experimental results show consistent improvements on both datasets over prior works and our baselines.

cs.CV↗

arXiv2019

Early Action Prediction with Generative Adversarial Networks

Action Prediction is aimed to determine what action is occurring in a video as early as possible, which is crucial to many online applications, such as predicting a traffic accident before it happens and detecting malicious actions in the monitoring system. In this work, we address this problem by developing an end-to-end architecture that improves the discriminability of features of partially observed videos by assimilating them to features from complete videos. For this purpose, the generative adversarial network is introduced for tackling action prediction problem, which improves the recognition accuracy of partially observed videos though narrowing the feature difference of partially observed videos from complete ones. Specifically, its generator comprises of two networks: a CNN for feature extraction and an LSTM for estimating residual error between features of the partially observed videos and complete ones, and then the features from CNN adds the residual error from LSTM, which is regarded as the enhanced feature to fool a competing discriminator. Meanwhile, the generator is trained with an additional perceptual objective, which forces the enhanced features of partially observed videos are discriminative enough for action prediction. Extensive experimental results on UCF101, BIT and UT-Interaction datasets demonstrate that our approach outperforms the state-of-the-art methods, especially for videos that less than 50% portion of frames is observed.

cs.CV↗

arXiv2019

Cross-Modal Message Passing for Two-stream Fusion

Processing and fusing information among multi-modal is a very useful technique for achieving high performance in many computer vision problems. In order to tackle multi-modal information more effectively, we introduce a novel framework for multi-modal fusion: Cross-modal Message Passing (CMMP). Specifically, we propose a cross-modal message passing mechanism to fuse two-stream network for action recognition, which composes of an appearance modal network (RGB image) and a motion modal (optical flow image) network. The objectives of individual networks in this framework are two-fold: a standard classification objective and a competing objective. The classification object ensures that each modal network predicts the true action category while the competing objective encourages each modal network to outperform the other one. We quantitatively show that the proposed CMMP fuses the traditional two-stream network more effectively, and outperforms all existing two-stream fusion method on UCF-101 and HMDB-51 datasets.

cs.CV↗

arXiv2019

Anomaly Detection in Traffic Scenes via Spatial-aware Motion Reconstruction

Anomaly detection from a driver's perspective when driving is important to autonomous vehicles. As a part of Advanced Driver Assistance Systems (ADAS), it can remind the driver about dangers timely. Compared with traditional studied scenes such as the university campus and market surveillance videos, it is difficult to detect abnormal event from a driver's perspective due to camera waggle, abidingly moving background, drastic change of vehicle velocity, etc. To tackle these specific problems, this paper proposes a spatial localization constrained sparse coding approach for anomaly detection in traffic scenes, which firstly measures the abnormality of motion orientation and magnitude respectively and then fuses these two aspects to obtain a robust detection result. The main contributions are threefold: 1) This work describes the motion orientation and magnitude of the object respectively in a new way, which is demonstrated to be better than the traditional motion descriptors. 2) The spatial localization of object is taken into account of the sparse reconstruction framework, which utilizes the scene's structural information and outperforms the conventional sparse coding methods. 3) Results of motion orientation and magnitude are adaptively weighted and fused by a Bayesian model, which makes the proposed method more robust and handle more kinds of abnormal events. The efficiency and effectiveness of the proposed method are validated by testing on nine difficult video sequences captured by ourselves. Observed from the experimental results, the proposed method is more effective and efficient than the popular competitors, and yields a higher performance.

cs.CV↗