arXiv subjects

Wei Wei

Publications and source records attributed to Wei Wei.

At least 451 records · Page 25Linked to original sources

arXiv2019

Learning with Hierarchical Complement Objective

Label hierarchies widely exist in many vision-related problems, ranging from explicit label hierarchies existed in image classification to latent label hierarchies existed in semantic segmentation. Nevertheless, state-of-the-art methods often deploy cross-entropy loss that implicitly assumes class labels to be exclusive and thus independence from each other. Motivated by the fact that classes from the same parental category usually share certain similarity, we design a new training diagram called Hierarchical Complement Objective Training (HCOT) that leverages the information from label hierarchy. HCOT maximizes the probability of the ground truth class, and at the same time, neutralizes the probabilities of rest of the classes in a hierarchical fashion, making the model take advantage of the label hierarchy explicitly. The proposed HCOT is evaluated on both image classification and semantic segmentation tasks. Experimental results confirm that HCOT outperforms state-of-the-art models in CIFAR-100, ImageNet-2012, and PASCAL-Context. The study further demonstrates that HCOT can be applied on tasks with latent label hierarchies, which is a common characteristic in many machine learning tasks.

cs.CV↗

arXiv2019

Meta Architecture Search

Neural Architecture Search (NAS) has been quite successful in constructing state-of-the-art models on a variety of tasks. Unfortunately, the computational cost can make it difficult to scale. In this paper, we make the first attempt to study Meta Architecture Search which aims at learning a task-agnostic representation that can be used to speed up the process of architecture search on a large number of tasks. We propose the Bayesian Meta Architecture SEarch (BASE) framework which takes advantage of a Bayesian formulation of the architecture search problem to learn over an entire set of tasks simultaneously. We show that on Imagenet classification, we can find a model that achieves 25.7% top-1 error and 8.1% top-5 error by adapting the architecture in less than an hour from an 8 GPU days pretrained meta-network. By learning a good prior for NAS, our method dramatically decreases the required computation cost while achieving comparable performance to current state-of-the-art methods - even finding competitive models for unseen datasets with very quick adaptation. We believe our framework will open up new possibilities for efficient and massively scalable architecture search research across multiple tasks.

cs.LG↗

arXiv2019

Table-to-Text Natural Language Generation with Unseen Schemas

Traditional table-to-text natural language generation (NLG) tasks focus on generating text from schemas that are already seen in the training set. This limitation curbs their generalizabilities towards real-world scenarios, where the schemas of input tables are potentially infinite. In this paper, we propose the new task of table-to-text NLG with unseen schemas, which specifically aims to test the generalization of NLG for input tables with attribute types that never appear during training. To do this, we construct a new benchmark dataset for this task. To deal with the problem of unseen attribute types, we propose a new model that first aligns unseen table schemas to seen ones, and then generates text with updated table representations. Experimental evaluation on the new benchmark demonstrates that our model outperforms baseline methods by a large margin. In addition, comparison with standard data-to-text settings shows the challenges and uniqueness of our proposed task.

cs.CL↗

arXiv2019

AIM 2019 Challenge on Constrained Super-Resolution: Methods and Results

This paper reviews the AIM 2019 challenge on constrained example-based single image super-resolution with focus on proposed solutions and results. The challenge had 3 tracks. Taking the three main aspects (i.e., number of parameters, inference/running time, fidelity (PSNR)) of MSRResNet as the baseline, Track 1 aims to reduce the amount of parameters while being constrained to maintain or improve the running time and the PSNR result, Tracks 2 and 3 aim to optimize running time and PSNR result with constrain of the other two aspects, respectively. Each track had an average of 64 registered participants, and 12 teams submitted the final results. They gauge the state-of-the-art in single image super-resolution.

eess.IV↗

arXiv2019

An Energy Sharing Game with Generalized Demand Bidding: Model and Properties

This paper proposes a novel energy sharing mechanism for prosumers who can produce and consume. Different from most existing works, the role of individual prosumer as a seller or buyer in our model is endogenously determined. Several desirable properties of the proposed mechanism are proved based on a generalized game-theoretic model. We show that the Nash equilibrium exists and is the unique solution of an equivalent convex optimization problem. The sharing price at the Nash equilibrium equals to the average marginal disutility of all prosumers. We also prove that every prosumer has the incentive to participate in the sharing market, and prosumers' total cost decreases with increasing absolute value of price sensitivity. Furthermore, the Nash equilibrium approaches the social optimal as the number of prosumers grows, and competition can improve social welfare.

math.OC↗

arXiv2019

$\varepsilon$-regularity criteria in Lorentz spaces to the 3D Navier-Stokes equations

In this paper, we are concerned with regularity of suitable weak solutions of the 3D Navier-Stokes equations in Lorentz spaces. We obtain $\varepsilon$-regularity criteria in terms of either the velocity, the gradient of the velocity, the vorticity, or deformation tensor in Lorentz spaces. As an application, this allows us to extend the result involving Leray's blow up rate in time, and to show that the number of singular points of weak solutions belonging to $ L^{p,\infty}(-1,0;L^{q,l}(\mathbb{R}^{3})) $ and $ {2}/{p}+{3}/{q}=1$ with $3<q<\infty$ and $q\leq l <\infty$ is finite.

math.AP↗

arXiv2019

AI Matrix: A Deep Learning Benchmark for Alibaba Data Centers

Alibaba has China's largest e-commerce platform. To support its diverse businesses, Alibaba has its own large-scale data centers providing the computing foundation for a wide variety of software applications. Among these applications, deep learning (DL) has been playing an important role in delivering services like image recognition, objection detection, text recognition, recommendation, and language processing. To build more efficient data centers that deliver higher performance for these DL applications, it is important to understand their computational needs and use that information to guide the design of future computing infrastructure. An effective way to achieve this is through benchmarks that can fully represent Alibaba's DL applications.

cs.LG↗

arXiv2019

New regularity criteria based on pressure or gradient of velocity in Lorentz spaces for the 3D Navier-Stokes equations

In this paper, we derive regular criteria via pressure or gradient of the velocity in Lorentz spaces to the 3D Navier-Stokes equations. It is shown that a Leray-Hopf weak solution is regular on $(0,T]$ provided that either the norm $\|Π\|_{L^{p,\infty}(0,T; L ^{q,\infty}(\mathbb{R}^{3}))} $ with $ {2}/{p}+{3}/{q}=2$ $({3}/{2}<q<\infty)$ or $\|\nablaΠ\|_{L^{p,\infty}(0,T; L ^{q,\infty}(\mathbb{R}^{3}))} $ with $ {2}/{p}+{3}/{q}=3$ $(1<q<\infty)$ is small. This gives an affirmative answer to a question proposed by Suzuki in [26, Remark 2.4, p.3850]. Moreover, regular conditions in terms of $\nabla u$ obtained here generalize known ones to allow the time direction to belong to Lorentz spaces.

math.AP↗

arXiv2019

Natural Adversarial Sentence Generation with Gradient-based Perturbation

This work proposes a novel algorithm to generate natural language adversarial input for text classification models, in order to investigate the robustness of these models. It involves applying gradient-based perturbation on the sentence embeddings that are used as the features for the classifier, and learning a decoder for generation. We employ this method to a sentiment analysis model and verify its effectiveness in inducing incorrect predictions by the model. We also conduct quantitative and qualitative analysis on these examples and demonstrate that our approach can generate more natural adversaries. In addition, it can be used to successfully perform black-box attacks, which involves attacking other existing models whose parameters are not known. On a public sentiment analysis API, the proposed method introduces a 20% relative decrease in average accuracy and 74% relative increase in absolute error.

cs.IR↗

arXiv2019

Stack-VS: Stacked Visual-Semantic Attention for Image Caption Generation

Recently, automatic image caption generation has been an important focus of the work on multimodal translation task. Existing approaches can be roughly categorized into two classes, i.e., top-down and bottom-up, the former transfers the image information (called as visual-level feature) directly into a caption, and the later uses the extracted words (called as semanticlevel attribute) to generate a description. However, previous methods either are typically based one-stage decoder or partially utilize part of visual-level or semantic-level information for image caption generation. In this paper, we address the problem and propose an innovative multi-stage architecture (called as Stack-VS) for rich fine-gained image caption generation, via combining bottom-up and top-down attention models to effectively handle both visual-level and semantic-level information of an input image. Specifically, we also propose a novel well-designed stack decoder model, which is constituted by a sequence of decoder cells, each of which contains two LSTM-layers work interactively to re-optimize attention weights on both visual-level feature vectors and semantic-level attribute embeddings for generating a fine-gained image caption. Extensive experiments on the popular benchmark dataset MSCOCO show the significant improvements on different evaluation metrics, i.e., the improvements on BLEU-4/CIDEr/SPICE scores are 0.372, 1.226 and 0.216, respectively, as compared to the state-of-the-arts.

cs.CV↗

arXiv2019

Failure of Smooth Pasting Principle and Nonexistence of Equilibrium Stopping Rules under Time-Inconsistency

This paper considers a time-inconsistent stopping problem in which the inconsistency arises from non-constant time preference rates. We show that the smooth pasting principle, the main approach that has been used to construct explicit solutions for conventional time-consistent optimal stopping problems, may fail under time-inconsistency. Specifically, we prove that the smooth pasting principle solves a time-inconsistent problem within the intra-personal game theoretic framework if and only if a certain inequality on the model primitives is satisfied. We show that the violation of this inequality can happen even for very simple non-exponential discount functions. Moreover, we demonstrate that the stopping problem does not admit any intra-personal equilibrium whenever the smooth pasting principle fails. The "negative" results in this paper caution blindly extending the classical approaches for time-consistent stopping problems to their time-inconsistent counterparts.

q-fin.MF↗

arXiv2019

Deep Hashing for Signed Social Network Embedding

Network embedding is a promising way of network representation, facilitating many signed social network processing and analysis tasks such as link prediction and node classification. Recently, feature hashing has been adopted in several existing embedding algorithms to improve the efficiency, which has obtained a great success. However, the existing feature hashing based embedding algorithms only consider the positive links in signed social networks. Intuitively, negative links can also help improve the performance. Thus, in this paper, we propose a novel deep hashing method for signed social network embedding by considering simultaneously positive and negative links. Extensive experiments show that the proposed method performs better than several state-of-the-art baselines through link prediction task over two real-world signed social networks.

cs.SI↗

arXiv2019

Enhanced 3D convolutional networks for crowd counting

Recently, convolutional neural networks (CNNs) are the leading defacto method for crowd counting. However, when dealing with video datasets, CNN-based methods still process each video frame independently, thus ignoring the powerful temporal information between consecutive frames. In this work, we propose a novel architecture termed as "temporal channel-aware" (TCA) block, which achieves the capability of exploiting the temporal interdependencies among video sequences. Specifically, we incorporate 3D convolution kernels to encode local spatio-temporal features. Furthermore, the global contextual information is encoded into modulation weights which adaptively recalibrate channel-aware feature responses. With the local and global context combined, the proposed block enhances the discriminative ability of the feature representations and contributes to more precise results in diverse scenes. By stacking TCA blocks together, we obtain the deep trainable architecture called enhanced 3D convolutional networks (E3D). The experiments on three benchmark datasets show that the proposed method delivers state-of-the-art performance. To verify the generality, an extended experiment is conducted on a vehicle dataset TRANCOS and our approach beats previous methods by large margins.

cs.CV↗

arXiv2019

Improving Adversarial Robustness via Guided Complement Entropy

Adversarial robustness has emerged as an important topic in deep learning as carefully crafted attack samples can significantly disturb the performance of a model. Many recent methods have proposed to improve adversarial robustness by utilizing adversarial training or model distillation, which adds additional procedures to model training. In this paper, we propose a new training paradigm called Guided Complement Entropy (GCE) that is capable of achieving "adversarial defense for free," which involves no additional procedures in the process of improving adversarial robustness. In addition to maximizing model probabilities on the ground-truth class like cross-entropy, we neutralize its probabilities on the incorrect classes along with a "guided" term to balance between these two terms. We show in the experiments that our method achieves better model robustness with even better performance compared to the commonly used cross-entropy training objective. We also show that our method can be used orthogonal to adversarial training across well-known methods with noticeable robustness gain. To the best of our knowledge, our approach is the first one that improves model robustness without compromising performance.

cs.LG↗

arXiv2019

Deep Cross-Modal Hashing with Hashing Functions and Unified Hash Codes Jointly Learning

Due to their high retrieval efficiency and low storage cost, cross-modal hashing methods have attracted considerable attention. Generally, compared with shallow cross-modal hashing methods, deep cross-modal hashing methods can achieve a more satisfactory performance by integrating feature learning and hash codes optimizing into a same framework. However, most existing deep cross-modal hashing methods either cannot learn a unified hash code for the two correlated data-points of different modalities in a database instance or cannot guide the learning of unified hash codes by the feedback of hashing function learning procedure, to enhance the retrieval accuracy. To address the issues above, in this paper, we propose a novel end-to-end Deep Cross-Modal Hashing with Hashing Functions and Unified Hash Codes Jointly Learning (DCHUC). Specifically, by an iterative optimization algorithm, DCHUC jointly learns unified hash codes for image-text pairs in a database and a pair of hash functions for unseen query image-text pairs. With the iterative optimization algorithm, the learned unified hash codes can be used to guide the hashing function learning procedure; Meanwhile, the learned hashing functions can feedback to guide the unified hash codes optimizing procedure. Extensive experiments on three public datasets demonstrate that the proposed method outperforms the state-of-the-art cross-modal hashing methods.

cs.IR↗

arXiv2019

Discriminative Embedding Autoencoder with a Regressor Feedback for Zero-Shot Learning

Zero-shot learning (ZSL) aims to recognize the novel object categories using the semantic representation of categories, and the key idea is to explore the knowledge of how the novel class is semantically related to the familiar classes. Some typical models are to learn the proper embedding between the image feature space and the semantic space, whilst it is important to learn discriminative features and comprise the coarse-to-fine image feature and semantic information. In this paper, we propose a discriminative embedding autoencoder with a regressor feedback model for ZSL. The encoder learns a mapping from the image feature space to the discriminative embedding space, which regulates both inter-class and intra-class distances between the learned features by a margin, making the learned features be discriminative for object recognition. The regressor feedback learns to map the reconstructed samples back to the the discriminative embedding and the semantic embedding, assisting the decoder to improve the quality of the samples and provide a generalization to the unseen classes. The proposed model is validated extensively on four benchmark datasets: SUN, CUB, AWA1, AWA2, the experiment results show that our proposed model outperforms the state-of-the-art models, and especially in the generalized zero-shot learning (GZSL), significant improvements are achieved.

cs.CV↗

arXiv2019

Buy or Sell? Energy Sharing of Prosumers on Constrained Networks

The advent of intelligent agents who produce and consume energy by themselves has led the smart grid into the era of "prosumer", offering the energy system and customers a unique opportunity to revaluate/trade their spot energy via a sharing initiative. To this end, designing an appropriate sharing mechanism is an issue with crucial importance and has captured great attention. This paper addresses the prosumers' demand response problem via energy sharing. Under a general supply-demand function bidding scheme, a sharing market clearing procedure considering network constraints is proposed, which gives rise to a generalized Nash game. The existence and uniqueness of market equilibrium are proved in non-congested cases. When congestion occurs, infinitely much equilibrium may exist because the strategy spaces of prosumers are correlated. A price-regulation procedure is introduced in the sharing mechanism, which outcomes a unique equilibrium that is fair to all participants. Properties of the improved sharing mechanism, including the individual rational behaviors of prosumers and the components of sharing price, are revealed. When the number of prosumers increases, the proposed sharing mechanism approaches social optimum. Even with fixed number of resources, introducing competition can result in a decreasing social cost. Illustrative examples validate the theoretical results and provide more insights for the energy sharing research.

math.OC↗

arXiv2019

Charged Higgs bosons in the NMSSM under current LHC constraints

Charged Higgs boson is a crucial prediction of new physics beyond the SM. In this work, we perform a comprehensive scan over the parameter space of NMSSM considering various experimental constraints including the direct search limits from the 13 TeV LHC, and consider the scenario that the next-to-lightest CP-even Higgs boson is SM-like. We find that the masses of charged Higgs bosons can be as light as 350 GeV, the lightest CP-even Higgs boson $h_1$ is predominantly singlet and can be as light as 48 GeV, and the lightest CP-odd Higgs boson $a_1$ is also singlet-dominated and can be as light as 82 GeV. The charged Higgs bosons mainly decay to $t\bar{b}$ or $\bar{t} b$, but the branching ratio of the exotic decays $H^\pm\to W^\pm h_1$ and $H^\pm\to W^\pm a_1$ can maximally reach to 20\% and 11\%, respectively, which can be used to distinguish the NMSSM from MSSM. Such a heavy charged Higgs boson is unaccessible at the 13 TeV LHC with a luminosity of 36.1 $\rm fb^{-1}$ and its detection needs higher energy and/or higher luminosity.

hep-ph↗