arXiv subjects

Dong Wang

Publications and source records attributed to Dong Wang.

At least 631 records · Page 35Linked to original sources

arXiv2020

Curriculum Audiovisual Learning

Associating sound and its producer in complex audiovisual scene is a challenging task, especially when we are lack of annotated training data. In this paper, we present a flexible audiovisual model that introduces a soft-clustering module as the audio and visual content detector, and regards the pervasive property of audiovisual concurrency as the latent supervision for inferring the correlation among detected contents. To ease the difficulty of audiovisual learning, we propose a novel curriculum learning strategy that trains the model from simple to complex scene. We show that such ordered learning procedure rewards the model the merits of easy training and fast convergence. Meanwhile, our audiovisual model can also provide effective unimodal representation and cross-modal alignment performance. We further deploy the well-trained model into practical audiovisual sound localization and separation task. We show that our localization model significantly outperforms existing methods, based on which we show comparable performance in sound separation without referring external visual supervision. Our video demo can be found at https://youtu.be/kuClfGG0cFU.

cs.CV↗

arXiv2020

Backreacting holographic superconductors from the coupling of a scalar field to the Einstein tensor

We investigate the properties of the backreacting holographic superconductors from the coupling of a scalar field to the Einstein tensor in the background of a d-dimensional AdS black hole. Imposing the Dirichlet boundary condition of the trial function without the Neumann boundary conditions, we improve the analytical Sturm-Liouville method with an iterative procedure to explore the pure effect of the Einstein tensor on the holographic superconductors and find that the Einstein tensor hinders the condensate of the scalar field but does not affect the critical phenomena. Our analytical findings are in very good agreement with the numerical results from the "marginally stable modes" method, which implies that the Sturm-Liouville method is still powerful to study the holographic superconductors from the coupling of a scalar field to the Einstein tensor even if we consider the backreactions.

gr-qc↗

arXiv2019

Experimental observation the Einstein-Podolsky-Rosen Steering based on the detection of entanglement

The Einstein-Podolsky-Rosen (EPR) steering is an intermediate quantum nonlocality between entanglement and Bell nonlocality, which plays an important role in quantum information processing tasks. In the past few years, the investigations concerning EPR steering have been demonstrated in a series of experiments. However, these studies rely on the relevant steering inequalities and the choices of measurement settings. Here, we experimentally verify the EPR steering via entanglement detection without using any steering inequality and measurement setting. By constructing two new states from a two-qubit target state, we observe the EPR steering by detecting the entanglement of these new states. The results show that the entanglement of the newly constructed states can be regarded as a new kind of steering witness for target states. Compared to the results of Xiao et al. [Phys. Rev. Lett. 118, 140404 (2017)], we find that the ability of detecting EPR steering in our scenario is stronger than two-setting projective measurements, which can observe more steerable states. Hence, our demonstrations can deepen the understanding of the connection between the EPR steering and entanglement.

quant-ph↗

arXiv2019

ICDAR 2019 Robust Reading Challenge on Reading Chinese Text on Signboard

Chinese scene text reading is one of the most challenging problems in computer vision and has attracted great interest. Different from English text, Chinese has more than 6000 commonly used characters and Chinesecharacters can be arranged in various layouts with numerous fonts. The Chinese signboards in street view are a good choice for Chinese scene text images since they have different backgrounds, fonts and layouts. We organized a competition called ICDAR2019-ReCTS, which mainly focuses on reading Chinese text on signboard. This report presents the final results of the competition. A large-scale dataset of 25,000 annotated signboard images, in which all the text lines and characters are annotated with locations and transcriptions, were released. Four tasks, namely character recognition, text line recognition, text line detection and end-to-end recognition were set up. Besides, considering the Chinese text ambiguity issue, we proposed a multi ground truth (multi-GT) evaluation method to make evaluation fairer. The competition started on March 1, 2019 and ended on April 30, 2019. 262 submissions from 46 teams are received. Most of the participants come from universities, research institutes, and tech companies in China. There are also some participants from the United States, Australia, Singapore, and Korea. 21 teams submit results for Task 1, 23 teams submit results for Task 2, 24 teams submit results for Task 3, and 13 teams submit results for Task 4. The official website for the competition is http://rrc.cvc.uab.es/?ch=12.

cs.CV↗

arXiv2019

Experimental investigation of entropic uncertainty relations and coherence uncertainty relations

Uncertainty relation usually is one of the most important features in quantum mechanics, and is the backbone of quantum theory, which distinguishes from the rule in classical counterpart. Specifically, entropy-based uncertainty relations are of fundamental importance in the region of quantum information theory, offering one nontrivial bound of key rate towards quantum key distribution. In this work, we experimentally demonstrate the entropic uncertainty relations and coherence-based uncertainty relations in an all-optics platform. By means of preparing two kinds of bipartite initial states with high fidelity, i.e., Bell-like states and Bell-like diagonal states, we carry on local projective measurements over a complete set of mutually unbiased bases on the measured subsystem. In terms of quantum tomography, the density matrices of the initial states and the post-measurement states are reconstructed. It shows that our experimental results coincide with the theoretical predictions very well. Additionally, we also verify that the lower bounds of both the entropy-based and coherence-based uncertainty can be tightened by imposing the Holevo quantity and mutual information, and the entropic uncertainty is inversely correlated with the coherence. Our demonstrations might offer an insight into their uncertainty relations and their connection to quantum coherence in quantum information science, which might be applicable to the security analysis of quantum key distributions.

quant-ph↗

arXiv2019

Generalized random matrix model with additional interactions

We introduce a log-gas model that is a generalization of a random matrix ensemble with an additional interaction, whose strength depends on a parameter $γ$. The equilibrium density is computed by numerically solving the Riemann-Hilbert problem associated with the ensemble. The effect of the additional parameter $γ$ associated with the two-body interaction can be understood in terms of an effective $γ$-dependent single-particle confining potential.

cond-mat.dis-nn↗

arXiv2019

Syntax-Infused Transformer and BERT models for Machine Translation and Natural Language Understanding

Attention-based models have shown significant improvement over traditional algorithms in several NLP tasks. The Transformer, for instance, is an illustrative example that generates abstract representations of tokens inputted to an encoder based on their relationships to all tokens in a sequence. Recent studies have shown that although such models are capable of learning syntactic features purely by seeing examples, explicitly feeding this information to deep learning models can significantly enhance their performance. Leveraging syntactic information like part of speech (POS) may be particularly beneficial in limited training data settings for complex models such as the Transformer. We show that the syntax-infused Transformer with multiple features achieves an improvement of 0.7 BLEU when trained on the full WMT 14 English to German translation dataset and a maximum improvement of 1.99 BLEU points when trained on a fraction of the dataset. In addition, we find that the incorporation of syntax into BERT fine-tuning outperforms baseline on a number of downstream tasks from the GLUE benchmark.

cs.CL↗

arXiv2019

ROI Pooled Correlation Filters for Visual Tracking

The ROI (region-of-interest) based pooling method performs pooling operations on the cropped ROI regions for various samples and has shown great success in the object detection methods. It compresses the model size while preserving the localization accuracy, thus it is useful in the visual tracking field. Though being effective, the ROI-based pooling operation is not yet considered in the correlation filter formula. In this paper, we propose a novel ROI pooled correlation filter (RPCF) algorithm for robust visual tracking. Through mathematical derivations, we show that the ROI-based pooling can be equivalently achieved by enforcing additional constraints on the learned filter weights, which makes the ROI-based pooling feasible on the virtual circular samples. Besides, we develop an efficient joint training formula for the proposed correlation filter algorithm, and derive the Fourier solvers for efficient model training. Finally, we evaluate our RPCF tracker on OTB-2013, OTB-2015 and VOT-2017 benchmark datasets. Experimental results show that our tracker performs favourably against other state-of-the-art trackers.

cs.CV↗

arXiv2019

CN-CELEB: a challenging Chinese speaker recognition dataset

Recently, researchers set an ambitious goal of conducting speaker recognition in unconstrained conditions where the variations on ambient, channel and emotion could be arbitrary. However, most publicly available datasets are collected under constrained environments, i.e., with little noise and limited channel variation. These datasets tend to deliver over optimistic performance and do not meet the request of research on speaker recognition in unconstrained conditions. In this paper, we present CN-Celeb, a large-scale speaker recognition dataset collected `in the wild'. This dataset contains more than 130,000 utterances from 1,000 Chinese celebrities, and covers 11 different genres in real world. Experiments conducted with two state-of-the-art speaker recognition approaches (i-vector and x-vector) show that the performance on CN-Celeb is far inferior to the one obtained on VoxCeleb, a widely used speaker recognition dataset. This result demonstrates that in real-life conditions, the performance of existing techniques might be much worse than it was thought. Our database is free for researchers and can be downloaded from http://project.cslt.org.

eess.AS↗

arXiv2019

On Investigation of Unsupervised Speech Factorization Based on Normalization Flow

Speech signals are complex composites of various information, including phonetic content, speaker traits, channel effect, etc. Decomposing this complicated mixture into independent factors, i.e., speech factorization, is fundamentally important and plays the central role in many important algorithms of modern speech processing tasks. In this paper, we present a preliminary investigation on unsupervised speech factorization based on the normalization flow model. This model constructs a complex invertible transform, by which we can project speech segments into a latent code space where the distribution is a simple diagonal Gaussian. Our preliminary investigation on the TIMIT database shows that this code space exhibits favorable properties such as denseness and pseudo linearity, and perceptually important factors such as phonetic content and speaker trait can be represented as particular directions within the code space.

cs.SD↗

arXiv2019

Gram-Gauss-Newton Method: Learning Overparameterized Neural Networks for Regression Problems

First-order methods such as stochastic gradient descent (SGD) are currently the standard algorithm for training deep neural networks. Second-order methods, despite their better convergence rate, are rarely used in practice due to the prohibitive computational cost in calculating the second-order information. In this paper, we propose a novel Gram-Gauss-Newton (GGN) algorithm to train deep neural networks for regression problems with square loss. Our method draws inspiration from the connection between neural network optimization and kernel regression of neural tangent kernel (NTK). Different from typical second-order methods that have heavy computational cost in each iteration, GGN only has minor overhead compared to first-order methods such as SGD. We also give theoretical results to show that for sufficiently wide neural networks, the convergence rate of GGN is \emph{quadratic}. Furthermore, we provide convergence guarantee for mini-batch GGN algorithm, which is, to our knowledge, the first convergence result for the mini-batch version of a second-order method on overparameterized neural networks. Preliminary experiments on regression tasks demonstrate that for training standard networks, our GGN algorithm converges much faster and achieves better performance than SGD.

cs.LG↗

arXiv2019

Experimental certification of steering criterion based on general entropic uncertainty relation

Quantum steering describes the phenomenon that one system can be immediately influenced by another with local measurements. It can be detected by the violation of a powerful and useful steering criterion from general entropic uncertainty relation. This criterion, in principle, can be evaluated straightforwardly and achieved by only probability distributions from a finite set of measurement settings. Herein, we experimentally verify the steering criterion by means of the two-photon Werner-like states and three Pauli measurements. The results indicate that quantum steering can be verified by the criterion in a convenient way. In particular, it is no need to perform the usual quantum state tomography in experiment, which reduces the required experimental resources greatly. Moreover, we demonstrate that the criterion is stronger than the linear one for the detecting quantum steering of the Werner-like states.

quant-ph↗

arXiv2019

GradNet: Gradient-Guided Network for Visual Object Tracking

The fully-convolutional siamese network based on template matching has shown great potentials in visual tracking. During testing, the template is fixed with the initial target feature and the performance totally relies on the general matching ability of the siamese network. However, this manner cannot capture the temporal variations of targets or background clutter. In this work, we propose a novel gradient-guided network to exploit the discriminative information in gradients and update the template in the siamese network through feed-forward and backward operations. Our algorithm performs feed-forward and backward operations to exploit the discriminative informaiton in gradients and capture the core attention of the target. To be specific, the algorithm can utilize the information from the gradient to update the template in the current frame. In addition, a template generalization training method is proposed to better use gradient information and avoid overfitting. To our knowledge, this work is the first attempt to exploit the information in the gradient for template update in siamese-based trackers. Extensive experiments on recent benchmarks demonstrate that our method achieves better performance than other state-of-the-art trackers.

cs.CV↗

arXiv2019

An Online Reinforcement Learning Approach to Quality-Cost-Aware Task Allocation for Multi-Attribute Social Sensing

Social sensing has emerged as a new sensing paradigm where humans (or devices on their behalf) collectively report measurements about the physical world. This paper focuses on a quality-cost-aware task allocation problem in multi-attribute social sensing applications. The goal is to identify a task allocation strategy (i.e., decide when and where to collect sensing data) to achieve an optimized tradeoff between the data quality and the sensing cost. While recent progress has been made to tackle similar problems, three important challenges have not been well addressed: (i) "online task allocation": the task allocation schemes need to respond quickly to the potentially large dynamics of the measured variables in social sensing; (ii) "multi-attribute constrained optimization": minimizing the overall sensing error given the dependencies and constraints of multiple attributes of the measured variables is a non-trivial problem to solve; (iii) "nonuniform task allocation cost": the task allocation cost in social sensing often has a nonuniform distribution which adds additional complexity to the optimized task allocation problem. This paper develops a Quality-Cost-Aware Online Task Allocation (QCO-TA) scheme to address the above challenges using a principled online reinforcement learning framework. We evaluate the QCO-TA scheme through a real-world social sensing application and the results show that our scheme significantly outperforms the state-of-the-art baselines in terms of both sensing accuracy and cost.

cs.LG↗

arXiv2019

Poster Abstract: A Dynamic Data-Driven Prediction Model for Sparse Collaborative Sensing Applications

A fundamental problem in collaborative sensing lies in providing an accurate prediction of critical events (e.g., hazardous environmental condition, urban abnormalities, economic trends). However, due to the resource constraints, collaborative sensing applications normally only collect measurements from a subset of physical locations and predict the measurements for the rest of locations. This problem is referred to as sparse collaborative sensing prediction. In this poster, we present a novel closed-loop prediction model by leveraging topic modeling and online learning techniques. We evaluate our scheme using a real-world collaborative sensing dataset. The initial results show that our proposed scheme outperforms the state-of-the-art baselines.

eess.SP↗

arXiv2019

'Skimming-Perusal' Tracking: A Framework for Real-Time and Robust Long-term Tracking

Compared with traditional short-term tracking, long-term tracking poses more challenges and is much closer to realistic applications. However, few works have been done and their performance have also been limited. In this work, we present a novel robust and real-time long-term tracking framework based on the proposed skimming and perusal modules. The perusal module consists of an effective bounding box regressor to generate a series of candidate proposals and a robust target verifier to infer the optimal candidate with its confidence score. Based on this score, our tracker determines whether the tracked object being present or absent, and then chooses the tracking strategies of local search or global search respectively in the next frame. To speed up the image-wide global search, a novel skimming module is designed to efficiently choose the most possible regions from a large number of sliding windows. Numerous experimental results on the VOT-2018 long-term and OxUvA long-term benchmarks demonstrate that the proposed method achieves the best performance and runs in real-time. The source codes are available at https://github.com/iiau-tracker/SPLT.

cs.CV↗

arXiv2019

AP19-OLR Challenge: Three Tasks and Their Baselines

This paper introduces the fourth oriental language recognition (OLR) challenge AP19-OLR, including the data profile, the tasks and the evaluation principles. The OLR challenge has been held successfully for three consecutive years, along with APSIPA Annual Summit and Conference (APSIPA ASC). The challenge this year still focuses on practical and challenging tasks, precisely (1) short-utterance LID, (2) cross-channel LID and (3) zero-resource LID. The event this year includes more languages and more real-life data provided by SpeechOcean and the NSFC M2ASR project. All the data is free for participants. Recipes for x-vector system and back-end evaluation are also conducted as baselines for the three tasks. The participants can refer to these online-published recipes to deploy LID systems for convenience. We report the baseline results on the three tasks and demonstrate that the three tasks are worth some efforts to achieve better performance.

eess.AS↗

arXiv2019

Inequality relation between entanglement and Bell nonlocality for arbitrary two-qubit states

Entanglement and Bell nonlocality are used to describe quantum inseparabilities. Bell-nonlocal states form a strict subset of entangled states. A natural question arises concerning how much territory Bell nonlocality occupies entanglement for a general two-qubit entangled state. In this work, we investigate the relation between entanglement and Bell nonlocality by using lots of randomly generated two-qubit states, and give out a constraint inequality relation between the two quantum resources. For studying the upper or lower boundary of the inequality relation, we discover maximally (minimally) nonlocal entangled states, which maximize (minimize) the value of the Bell nonlocality for a given value of the entanglement. Futhermore, we consider a special kind of mixed state transformed by performing an arbitrary unitary operation on werner state. It is found that the special mixed state's entanglement and Bell nonlocality are related to ones of a pure state transformed by the unitary operation performed on the Bell state.

quant-ph↗