arXiv subjects

Kai Wang

Publications and source records attributed to Kai Wang.

At least 685 records · Page 38Linked to original sources

arXiv2021

A Benchmark of Ocular Disease Intelligent Recognition: One Shot for Multi-disease Detection

In ophthalmology, early fundus screening is an economic and effective way to prevent blindness caused by ophthalmic diseases. Clinically, due to the lack of medical resources, manual diagnosis is time-consuming and may delay the condition. With the development of deep learning, some researches on ophthalmic diseases have achieved good results, however, most of them are just based on one disease. During fundus screening, ophthalmologists usually give diagnoses of multi-disease on binocular fundus image, so we release a dataset with 8 diseases to meet the real medical scene, which contains 10,000 fundus images from both eyes of 5,000 patients. We did some benchmark experiments on it through some state-of-the-art deep neural networks. We found simply increasing the scale of network cannot bring good results for multi-disease classification, and a well-structured feature fusion method combines characteristics of multi-disease is needed. Through this work, we hope to advance the research of related fields.

cs.CV↗

arXiv2021

Applications of Deep Learning in Fundus Images: A Review

The use of fundus images for the early screening of eye diseases is of great clinical importance. Due to its powerful performance, deep learning is becoming more and more popular in related applications, such as lesion segmentation, biomarkers segmentation, disease diagnosis and image synthesis. Therefore, it is very necessary to summarize the recent developments in deep learning for fundus images with a review paper. In this review, we introduce 143 application papers with a carefully designed hierarchy. Moreover, 33 publicly available datasets are presented. Summaries and analyses are provided for each task. Finally, limitations common to all tasks are revealed and possible solutions are given. We will also release and regularly update the state-of-the-art results and newly-released datasets at https://github.com/nkicsl/Fundus Review to adapt to the rapid development of this field.

eess.IV↗

arXiv2021

Gaussian limit for determinantal point processes with $J$-Hermitian kernels

We show that the central limit theorem for linear statistics over determinantal point processes with $J$-Hermitian kernels holds under fairly general conditions. In particular, We establish Gaussian limit for linear statistics over determinantal point processes on union of two copies of $\mathbb{R}^d$ when the correlation kernels are $J$-Hermitian translation-invariant.

math.PR↗

arXiv2020

Exploring Emotion Features and Fusion Strategies for Audio-Video Emotion Recognition

The audio-video based emotion recognition aims to classify a given video into basic emotions. In this paper, we describe our approaches in EmotiW 2019, which mainly explores emotion features and feature fusion strategies for audio and visual modality. For emotion features, we explore audio feature with both speech-spectrogram and Log Mel-spectrogram and evaluate several facial features with different CNN models and different emotion pretrained strategies. For fusion strategies, we explore intra-modal and cross-modal fusion methods, such as designing attention mechanisms to highlights important emotion feature, exploring feature concatenation and factorized bilinear pooling (FBP) for cross-modal feature fusion. With careful evaluation, we obtain 65.5% on the AFEW validation set and 62.48% on the test set and rank third in the challenge.

cs.CV↗

arXiv2020

Modulating nano-inhomogeneity at electrode-solid electrolyte interfaces for dendrite-proof solid-state batteries and long-life memristors

The penetration of dendrites in ceramic lithium conductors severely constrains the development of solid-state batteries (SSBs) while its nanoscopic origin remain unelucidated. We develop an in-situ nanoscale electrochemical characterization technique to reveal the nanoscopic lithium dendrite growth kinetics and use it as a guiding tool to unlock the design of interfaces for dendrite-proof SSBs. Using Li7La3Zr2O12 (LLZO) as a model system, in-situ nanoscopic dendrite triggering measurements, ex-situ electro-mechanical characterizations, and finite element simulations are carried out which reveal the dominating role of Li+ flux detouring and nano-mechanical inhomogeneity on dendrite penetration. To mitigate such nano-inhomogeneity, an ionic-conductive homogenizing layer based on poly(propylene carbonate) is designed which in-situ reacts with lithium to form a highly conformal interphase at mild conditions. A high critical current density of 1.8mA cm-2 and a low interfacial resistance of 14Ω cm2 is achieved. Practical SSBs based on LiFePO4 cathodes show great cyclic stability without capacity decay over 300 cycles. Beyond this, highly reversible electrochemical dendrite healing behavior in LLZO is discovered using the nano-electrode, based on which a model memristor with a high on/off ratio of ~10^5 is demonstrated for >200 cycles. This work not only provides a novel tool to investigate and design interfaces in SSBs but offers also new opportunities for solid electrolytes beyond energy applications.

physics.app-ph↗

arXiv2020

PA-Cache: Evolving Learning-Based Popularity-Aware Content Caching in Edge Networks

As ubiquitous and personalized services are growing boomingly, an increasingly large amount of traffic is generated over the network by massive mobile devices. As a result, content caching is gradually extending to network edges to provide low-latency services, improve quality of service, and reduce redundant data traffic. Compared to the conventional content delivery networks, caches in edge networks with smaller sizes usually have to accommodate more bursty requests. In this paper, we propose an evolving learning-based content caching policy, named PA-Cache in edge networks. It adaptively learns time-varying content popularity and determines which contents should be replaced when the cache is full. Unlike conventional deep neural networks (DNNs), which learn a fine-tuned but possibly outdated or biased prediction model using the entire training dataset with high computational complexity, PA-Cache weighs a large set of content features and trains the multi-layer recurrent neural network from shallow to deeper when more requests arrive over time. We extensively evaluate the performance of our proposed PA-Cache on real-world traces from a large online video-on-demand service provider. \rb{The results show that PA-Cache outperforms existing popular caching algorithms and approximates the optimal algorithm with only a 3.8\% performance gap when the cache percentage is 1.0\%}. PA-Cache also significantly reduces the computational cost compared to conventional DNN-based approaches.

cs.NI↗

arXiv2020

PiPoMonitor: Mitigating Cross-core Cache Attacks Using the Auto-Cuckoo Filter

Cache side channel attacks obtain victim cache line access footprint to infer security-critical information. Among them, cross-core attacks exploiting the shared last level cache are more threatening as their simplicity to set up and high capacity. Stateful approaches of detection-based mitigation observe precise cache behaviors and protect specific cache lines that are suspected of being attacked. However, their recording structures incur large storage overhead and are vulnerable to reverse engineering attacks. Exploring the intrinsic non-determinate layout of a traditional Cuckoo filter, this paper proposes a space efficient Auto-Cuckoo filter to record access footprints, which succeed to decrease storage overhead and resist reverse engineering attacks at the same time. With Auto-Cuckoo filter, we propose PiPoMonitor to detect \textit{Ping-Pong patterns} and prefetch specific cache line to interfere with adversaries' cache probes. Security analysis shows the PiPoMonitor can effectively mitigate cross-core attacks and the Auto-Cuckoo filter is immune to reverse engineering attacks. Evaluation results indicate PiPoMonitor has negligible impact on performance and the storage overhead is only 0.37$\%$, an order of magnitude lower than previous stateful approaches.

cs.CR↗

arXiv2020

Efficient and Effective Community Search on Large-scale Bipartite Graphs

Bipartite graphs are widely used to model relationships between two types of entities. Community search retrieves densely connected subgraphs containing a query vertex, which has been extensively studied on unipartite graphs. However, community search on bipartite graphs remains largely unexplored. Moreover, all existing cohesive subgraph models on bipartite graphs can only be applied to measure the structure cohesiveness between two sets of vertices while overlooking the edge weight in forming the community. In this paper, we study the significant (alpha, beta)-community search problem on weighted bipartite graphs. Given a query vertex q, we aim to find the significant (alpha, beta)-community R of q which adopts (alpha, beta)-core to characterize the engagement level of vertices, and maximizes the minimum edge weight (significance) within R. To support fast retrieval of R, we first retrieve the maximal connected subgraph of (alpha, beta)-core containing the query vertex (the (alpha, beta)-community), and the search space is limited to this subgraph with a much smaller size than the original graph. A novel index structure is presented which can be built in O(delta * m) time and takes O(delta * m) space where m is the number of edges in G, delta is bounded by the square root of m and is much smaller in practice. Utilizing the index, the (alpha, beta)-community can be retrieved in optimal time. To further obtain R, we develop peeling and expansion algorithms to conduct searches by shrinking from the (alpha, beta)-community and expanding from the query vertex, respectively. The experimental results on real graphs not only demonstrate the effectiveness of the significant (alpha, beta)-community model but also validate the efficiency of our query processing and indexing techniques.

cs.DB↗

arXiv2020

Monitoring AGNs with Hβ Asymmetry. II. Reverberation Mapping of Three Seyfert Galaxies Historically Displaying Hβ Profiles with Changing Asymmetry: Mrk 79, NGC 3227, and Mrk 841

We report the results of reverberation mapping three bright Seyfert galaxies, Mrk 79, NGC 3227, and Mrk 841, from a campaign conducted from December 2016 to May 2017 with the Wyoming Infrared Observatory (WIRO) 2.3-meter telescope. All three of these targets have shown asymmetric broad H$β$ emission lines in the past, although their emission lines were relatively symmetric during our observations. We measured Hβ time lags for all three targets and estimated masses of their black holes -- for the first time in the case of Mrk 841. For Mrk 79 and NGC 3227, the data are of sufficient quality to resolve distinct time lags as a function of velocity and to compute two-dimensional velocity-delay maps. Mrk 79 shows smaller time lags for high-velocity gas but the distribution is not symmetric, and its complex velocity-delay map could result from the combination of both inflowing and outflowing Hβ emitting disks that may be part of a single larger structure. NGC 3227 shows the largest time lags for blueshifted gas and the two-dimensional velocity-delay map suggests a disk with some inflow. We compare our results with previous work and find evidence for different time lags despite similar luminosities, as well as evolving broad line region structures.

astro-ph.GA↗

arXiv2020

Suppressing Mislabeled Data via Grouping and Self-Attention

Deep networks achieve excellent results on large-scale clean data but degrade significantly when learning from noisy labels. To suppressing the impact of mislabeled data, this paper proposes a conceptually simple yet efficient training block, termed as Attentive Feature Mixup (AFM), which allows paying more attention to clean samples and less to mislabeled ones via sample interactions in small groups. Specifically, this plug-and-play AFM first leverages a \textit{group-to-attend} module to construct groups and assign attention weights for group-wise samples, and then uses a \textit{mixup} module with the attention weights to interpolate massive noisy-suppressed samples. The AFM has several appealing benefits for noise-robust deep learning. (i) It does not rely on any assumptions and extra clean subset. (ii) With massive interpolations, the ratio of useless samples is reduced dramatically compared to the original noisy ratio. (iii) \pxj{It jointly optimizes the interpolation weights with classifiers, suppressing the influence of mislabeled data via low attention weights. (iv) It partially inherits the vicinal risk minimization of mixup to alleviate over-fitting while improves it by sampling fewer feature-target vectors around mislabeled data from the mixup vicinal distribution.} Extensive experiments demonstrate that AFM yields state-of-the-art results on two challenging real-world noisy datasets: Food101N and Clothing1M. The code will be available at https://github.com/kaiwang960112/AFM.

cs.CV↗

arXiv2020

Vision-based Robotic Grasping From Object Localization, Object Pose Estimation to Grasp Estimation for Parallel Grippers: A Review

This paper presents a comprehensive survey on vision-based robotic grasping. We conclude three key tasks during vision-based robotic grasping, which are object localization, object pose estimation and grasp estimation. In detail, the object localization task contains object localization without classification, object detection and object instance segmentation. This task provides the regions of the target object in the input data. The object pose estimation task mainly refers to estimating the 6D object pose and includes correspondence-based methods, template-based methods and voting-based methods, which affords the generation of grasp poses for known objects. The grasp estimation task includes 2D planar grasp methods and 6DoF grasp methods, where the former is constrained to grasp from one direction. These three tasks could accomplish the robotic grasping with different combinations. Lots of object pose estimation methods need not object localization, and they conduct object localization and object pose estimation jointly. Lots of grasp estimation methods need not object localization and object pose estimation, and they conduct grasp estimation in an end-to-end manner. Both traditional methods and latest deep learning-based methods based on the RGB-D image inputs are reviewed elaborately in this survey. Related datasets and comparisons between state-of-the-art methods are summarized as well. In addition, challenges about vision-based robotic grasping and future directions in addressing these challenges are also pointed out.

cs.RO↗

arXiv2020

Automatically Learning Compact Quality-aware Surrogates for Optimization Problems

Solving optimization problems with unknown parameters often requires learning a predictive model to predict the values of the unknown parameters and then solving the problem using these values. Recent work has shown that including the optimization problem as a layer in the model training pipeline results in predictions of the unobserved parameters that lead to higher decision quality. Unfortunately, this process comes at a large computational cost because the optimization problem must be solved and differentiated through in each training iteration; furthermore, it may also sometimes fail to improve solution quality due to non-smoothness issues that arise when training through a complex optimization layer. To address these shortcomings, we learn a low-dimensional surrogate model of a large optimization problem by representing the feasible space in terms of meta-variables, each of which is a linear combination of the original variables. By training a low-dimensional surrogate model end-to-end, and jointly with the predictive model, we achieve: i) a large reduction in training and inference time; and ii) improved performance by focusing attention on the more important variables in the optimization and learning in a smoother space. Empirically, we demonstrate these improvements on a non-convex adversary modeling task, a submodular recommendation task and a convex portfolio optimization task.

cs.LG↗

arXiv2020

Cohomology theory of averaging algebras, $L_\infty$-structures and homotopy averaging algebras

This paper studies averaging algebras, say, associative algebras endowed with averaging operators. We develop a cohomology theory for averaging algebras and justify it by interpreting lower degree cohomology groups as formal deformations and abelian extensions of averaging algebras. We make explicit the $L_\infty$-algebra structure over the cochain complex defining cohomology groups and introduce the notion of homotopy averaging algebras as Maurer-Cartan elements of this $L_\infty$-algebra.

math.KT↗

arXiv2020

ShapeAssembly: Learning to Generate Programs for 3D Shape Structure Synthesis

Manually authoring 3D shapes is difficult and time consuming; generative models of 3D shapes offer compelling alternatives. Procedural representations are one such possibility: they offer high-quality and editable results but are difficult to author and often produce outputs with limited diversity. On the other extreme are deep generative models: given enough data, they can learn to generate any class of shape but their outputs have artifacts and the representation is not editable. In this paper, we take a step towards achieving the best of both worlds for novel 3D shape synthesis. We propose ShapeAssembly, a domain-specific "assembly-language" for 3D shape structures. ShapeAssembly programs construct shapes by declaring cuboid part proxies and attaching them to one another, in a hierarchical and symmetrical fashion. Its functions are parameterized with free variables, so that one program structure is able to capture a family of related shapes. We show how to extract ShapeAssembly programs from existing shape structures in the PartNet dataset. Then we train a deep generative model, a hierarchical sequence VAE, that learns to write novel ShapeAssembly programs. The program captures the subset of variability that is interpretable and editable. The deep model captures correlations across shape collections that are hard to express procedurally. We evaluate our approach by comparing shapes output by our generated programs to those from other recent shape structure synthesis models. We find that our generated shapes are more plausible and physically-valid than those of other methods. Additionally, we assess the latent spaces of these models, and find that ours is better structured and produces smoother interpolations. As an application, we use our generative model and differentiable program interpreter to infer and fit shape programs to unstructured geometry, such as point clouds.

cs.GR↗

arXiv2020

Identifying galaxy groups at high redshift from incomplete spectroscopic data: I. The group finder and application to zCOSMOS

Identifying galaxy groups from redshift surveys of galaxies plays an important role in connecting galaxies with the underlying dark matter distribution. Current and future high-$z$ spectroscopic surveys, usually incomplete in redshift sampling, present both opportunities and challenges to identifying groups in the high-$z$ Universe. We develop a group finder that is based on incomplete redshift samples combined with photometric data, using a machine learning method to assign halo masses to identified groups. Test using realistic mock catalogs shows that $\gtrsim 90\%$ of true groups with halo masses $\rm M_h \gtrsim 10^{12} M_{\odot}/h$ are successfully identified, and that the fraction of contaminants is smaller than $10\%$. The standard deviation in the halo mass estimation is smaller than 0.25 dex at all masses. We apply our group finder to zCOSMOS-bright and describe basic properties of the group catalog obtained.

astro-ph.GA↗

arXiv2020

Limiting Superluminal Neutrino Velocity and Lorentz Invariance Violation by Neutrino Emission from the Blazar TXS 0506+056

The detection of high-energy neutrino coincident with the blazar TXS 0506+056 provides a unique opportunity to test Lorentz invariance violation (LIV) in the neutrino sector. Thanks to the precisely measured redshift, i.e., $z=0.3365$, the comoving distance of the neutrino source is determined. In this work, we obtain and discuss the constraints on the superluminal neutrino velocity $δ_ν$ and the LIV by considering the energy loss of superluminal neutrino during propagation. Given superluminal electron velocity ($δ_e \ge 0$), a very stringent constraint on superluminal neutrino velocity can be reached, i.e., $δ_ν\lesssim 1.3\times 10^{-18}$, corresponding to the quantum gravity (QG) scale $M_{\rm QG,1} \gtrsim 5.7 \times 10^{3} M_{\rm Pl}$ and $M_{\rm QG,2} \gtrsim 9.3 \times 10^{-6} M_{\rm Pl}$ for linear (quadratic) LIV, which are $\sim 12$ orders of magnitude tighter for linear LIV and $\sim 9$ orders tighter for quadratic LIV compared to the time-of-flight constraint from MeV neutrinos of SN 1987A. While given the subluminal electron velocity, a weaker constraint on the superluminal neutrino velocity is obtained, i.e., $δ_ν\lesssim 8 \times 10^{-17}$, which is consistent with the conclusions of previous works. We also study the neutrino detection probability due to the distortion of neutrino spectral shape during propagation, which gives slightly weaker constraints than above by a factor of $\sim2$.

astro-ph.HE↗

arXiv2020

Constraining the Jet Launching Time of GRB 170817A by Utilizing the Baryon Loading

The observed delay of GRB 170817A relative to GW170817 carries significant information about gamma-ray burst (GRB) physics and is subject to intense debate. In this letter, we put forward an approach to discuss the major source of this time delay. First of all, we use the structured jet model to fit the X-ray/optical/radio afterglows of GRB 170817A together with superluminal motion measured by the Very Long Baseline Interferometry. Our structured jet is modelled with angle-dependent energy and baryon loading. It is found that our model can well fit the afterglows of GRB 170817A. After that, the baryon loading in the jet is inferred based on our fitting results. By comparing the baryon loading to the mass outflow in different stages, we infer that the time lag of the jet launch relative to the merger is less than hundreds or tens of milliseconds. It suggests that the time delay of GRB 170817A relative to GW170817 is defined mostly by the spreading time of the jet propagating to its dissipation radius.

astro-ph.HE↗

arXiv2020

Bookworm continual learning: beyond zero-shot learning and continual learning

We propose bookworm continual learning(BCL), a flexible setting where unseen classes can be inferred via a semantic model, and the visual model can be updated continually. Thus BCL generalizes both continual learning (CL) and zero-shot learning (ZSL). We also propose the bidirectional imagination (BImag) framework to address BCL where features of both past and future classes are generated. We observe that conditioning the feature generator on attributes can actually harm the continual learning ability, and propose two variants (joint class-attribute conditioning and asymmetric generation) to alleviate this problem.

cs.CV↗