arXiv subjects

Hao Yang

Publications and source records attributed to Hao Yang.

At least 415 records · Page 23Linked to original sources

arXiv2019

Position Focused Attention Network for Image-Text Matching

Image-text matching tasks have recently attracted a lot of attention in the computer vision field. The key point of this cross-domain problem is how to accurately measure the similarity between the visual and the textual contents, which demands a fine understanding of both modalities. In this paper, we propose a novel position focused attention network (PFAN) to investigate the relation between the visual and the textual views. In this work, we integrate the object position clue to enhance the visual-text joint-embedding learning. We first split the images into blocks, by which we infer the relative position of region in the image. Then, an attention mechanism is proposed to model the relations between the image region and blocks and generate the valuable position feature, which will be further utilized to enhance the region expression and model a more reliable relationship between the visual image and the textual sentence. Experiments on the popular datasets Flickr30K and MS-COCO show the effectiveness of the proposed method. Besides the public datasets, we also conduct experiments on our collected practical large-scale news dataset (Tencent-News) to validate the practical application value of proposed method. As far as we know, this is the first attempt to test the performance on the practical application. Our method achieves the state-of-art performance on all of these three datasets.

cs.CL↗

arXiv2019

CLCI-Net: Cross-Level fusion and Context Inference Networks for Lesion Segmentation of Chronic Stroke

Segmenting stroke lesions from T1-weighted MR images is of great value for large-scale stroke rehabilitation neuroimaging analyses. Nevertheless, there are great challenges with this task, such as large range of stroke lesion scales and the tissue intensity similarity. The famous encoder-decoder convolutional neural network, which although has made great achievements in medical image segmentation areas, may fail to address these challenges due to the insufficient uses of multi-scale features and context information. To address these challenges, this paper proposes a Cross-Level fusion and Context Inference Network (CLCI-Net) for the chronic stroke lesion segmentation from T1-weighted MR images. Specifically, a Cross-Level feature Fusion (CLF) strategy was developed to make full use of different scale features across different levels; Extending Atrous Spatial Pyramid Pooling (ASPP) with CLF, we have enriched multi-scale features to handle the different lesion sizes; In addition, convolutional long short-term memory (ConvLSTM) is employed to infer context information and thus capture fine structures to address the intensity similarity issue. The proposed approach was evaluated on an open-source dataset, the Anatomical Tracings of Lesions After Stroke (ATLAS) with the results showing that our network outperforms five state-of-the-art methods. We make our code and models available at https://github.com/YH0517/CLCI_Net.

eess.IV↗

arXiv2019

Predicting Next-Season Designs on High Fashion Runway

Fashion is a large and fast-changing industry. Foreseeing the upcoming fashion trends is beneficial for fashion designers, consumers, and retailers. However, fashion trends are often perceived as unpredictable due to the enormous amount of factors involved into designers' subjectivity. In this paper, we propose a fashion trend prediction framework and design neural network models to leverage structured fashion runway show data, learn the fashion collection embedding, and further train RNN/LSTM models to capture the designers' style evolution. Our proposed framework consists of (1) a runway embedding learning model that uses fashion runway images to learn every season's collection embedding, and (2) a next-season fashion design prediction model that leverage the concept of designer style and trend to predict next-season design given designers. Through experiments on a collected dataset across 32 years of fashion shows, our framework can achieve the best performance of 78.42% AUC on average and 95% for an individual designer when predicting the next season's design.

cs.CV↗

arXiv2019

Dynamic Graph Representation Learning via Self-Attention Networks

Learning latent representations of nodes in graphs is an important and ubiquitous task with widespread applications such as link prediction, node classification, and graph visualization. Previous methods on graph representation learning mainly focus on static graphs, however, many real-world graphs are dynamic and evolve over time. In this paper, we present Dynamic Self-Attention Network (DySAT), a novel neural architecture that operates on dynamic graphs and learns node representations that capture both structural properties and temporal evolutionary patterns. Specifically, DySAT computes node representations by jointly employing self-attention layers along two dimensions: structural neighborhood and temporal dynamics. We conduct link prediction experiments on two classes of graphs: communication networks and bipartite rating networks. Our experimental results show that DySAT has a significant performance gain over several different state-of-the-art graph embedding baselines.

cs.LG↗

arXiv2019

Face Parsing with RoI Tanh-Warping

Face parsing computes pixel-wise label maps for different semantic components (e.g., hair, mouth, eyes) from face images. Existing face parsing literature have illustrated significant advantages by focusing on individual regions of interest (RoIs) for faces and facial components. However, the traditional crop-and-resize focusing mechanism ignores all contextual area outside the RoIs, and thus is not suitable when the component area is unpredictable, e.g. hair. Inspired by the physiological vision system of human, we propose a novel RoI Tanh-warping operator that combines the central vision and the peripheral vision together. It addresses the dilemma between a limited sized RoI for focusing and an unpredictable area of surrounding context for peripheral information. To this end, we propose a novel hybrid convolutional neural network for face parsing. It uses hierarchical local based method for inner facial components and global methods for outer facial components. The whole framework is simple and principled, and can be trained end-to-end. To facilitate future research of face parsing, we also manually relabel the training data of the HELEN dataset and will make it public. Experiments on both HELEN and LFW-PL benchmarks demonstrate that our method surpasses state-of-the-art methods.

cs.CV↗

arXiv2019

The Pupil Has Become the Master: Teacher-Student Model-Based Word Embedding Distillation with Ensemble Learning

Recent advances in deep learning have facilitated the demand of neural models for real applications. In practice, these applications often need to be deployed with limited resources while keeping high accuracy. This paper touches the core of neural models in NLP, word embeddings, and presents a new embedding distillation framework that remarkably reduces the dimension of word embeddings without compromising accuracy. A novel distillation ensemble approach is also proposed that trains a high-efficient student model using multiple teacher models. In our approach, the teacher models play roles only during training such that the student model operates on its own without getting supports from the teacher models during decoding, which makes it eighty times faster and lighter than other typical ensemble methods. All models are evaluated on seven document classification datasets and show a significant advantage over the teacher models for most cases. Our analysis depicts insightful transformation of word embeddings from distillation and suggests a future direction to ensemble approaches using neural models.

cs.IR↗

arXiv2019

Mask-Guided Portrait Editing with Conditional GANs

Portrait editing is a popular subject in photo manipulation. The Generative Adversarial Network (GAN) advances the generating of realistic faces and allows more face editing. In this paper, we argue about three issues in existing techniques: diversity, quality, and controllability for portrait synthesis and editing. To address these issues, we propose a novel end-to-end learning framework that leverages conditional GANs guided by provided face masks for generating faces. The framework learns feature embeddings for every face component (e.g., mouth, hair, eye), separately, contributing to better correspondences for image translation, and local face editing. With the mask, our network is available to many applications, like face synthesis driven by mask, face Swap+ (including hair in swapping), and local manipulation. It can also boost the performance of face parsing a bit as an option of data augmentation.

cs.CV↗

arXiv2019

Exploring Bosonic Mediator of Interaction at BESIII

We present a comprehensive investigation on the possibility of the search for new force mediator $X$ boson in $e^+e^-$ collision and $J/ψ$ decay at the BESIII experiment. The typical interactions of $X$ boson coupling to leptons and quarks are explored. The production and decay properties of this $X$ particle, the product/decay chains $e^+e^-\to Xγ\to e^+e^-γ$ and $J/ψ\to Xγ\toμ^+μ^-γ$, and exclusion limits on the reduced coupling strength parameters as functions of $X$ boson mass are presented. With the data set of tens of fb$^{-1}~e^+e^-$ or $10^{10}~J/ψ$, we find that the exclusion limits on the coupling strength parameters fall in the range of $10^{-3}\sim10^{-4}$, depending on $m_X$ assuming the decay width 10 eV$<Γ_X<$100 eV reasonably, for various hypotheses in the literature. According to our estimation, the search for new force mediator $X$ boson in both $e^+e^-$ collision and $J/ψ$ decay are accessible in nowadays BESIII experiment.

hep-ph↗

arXiv2019

Real-Time Steganalysis for Stream Media Based on Multi-channel Convolutional Sliding Windows

Previous VoIP steganalysis methods face great challenges in detecting speech signals at low embedding rates, and they are also generally difficult to perform real-time detection, making them hard to truly maintain cyberspace security. To solve these two challenges, in this paper, combined with the sliding window detection algorithm and Convolution Neural Network we propose a real-time VoIP steganalysis method which based on multi-channel convolution sliding windows. In order to analyze the correlations between frames and different neighborhood frames in a VoIP signal, we define multi channel sliding detection windows. Within each sliding window, we design two feature extraction channels which contain multiple convolution layers with multiple convolution kernels each layer to extract correlation features of the input signal. Then based on these extracted features, we use a forward fully connected network for feature fusion. Finally, by analyzing the statistical distribution of these features, the discriminator will determine whether the input speech signal contains covert information or not.We designed several experiments to test the proposed model's detection ability under various conditions, including different embedding rates, different speech length, etc. Experimental results showed that the proposed model outperforms all the previous methods, especially in the case of low embedding rate, which showed state-of-the-art performance. In addition, we also tested the detection efficiency of the proposed model, and the results showed that it can achieve almost real-time detection of VoIP speech signals.

cs.MM↗

arXiv2018

Zero-Annotation Object Detection with Web Knowledge Transfer

Object detection is one of the major problems in computer vision, and has been extensively studied. Most of the existing detection works rely on labor-intensive supervision, such as ground truth bounding boxes of objects or at least image-level annotations. On the contrary, we propose an object detection method that does not require any form of human annotation on target tasks, by exploiting freely available web images. In order to facilitate effective knowledge transfer from web images, we introduce a multi-instance multi-label domain adaption learning framework with two key innovations. First of all, we propose an instance-level adversarial domain adaptation network with attention on foreground objects to transfer the object appearances from web domain to target domain. Second, to preserve the class-specific semantic structure of transferred object features, we propose a simultaneous transfer mechanism to transfer the supervision across domains through pseudo strong label generation. With our end-to-end framework that simultaneously learns a weakly supervised detector and transfers knowledge across domains, we achieved significant improvements over baseline methods on the benchmark datasets.

cs.CV↗

arXiv2017

Symmetry Demanded Topological Nodal-line Materials

The realization of Dirac and Weyl physics in solids has made topological materials one of the main focuses of condensed matter physics. Recently, the topic of topological nodal line semimetals, materials in which Dirac or Weyl-like crossings along special lines in momentum space create either a closed ring or line of degeneracies, rather than discrete points, has become a hot topic in topological quantum matter. Here we review the experimentally confirmed and theoretically predicted topological nodal line semimetals, focusing in particular on the symmetry protection mechanisms of the nodal lines in various materials. Three different mechanisms: a combination of inversion and time-reversal symmetry, mirror reflection symmetry, and non-symmorphic symmetry, and their robustness under the effect of spin orbit coupling are discussed. We also present a new Weyl nodal line material, the Te-square net compound KCu$_2$EuTe$_4$, which has several Weyl nodal lines including one extremely close to the Fermi level ($<$30 meV below E$_F$). Finally, we discuss potential experimental signatures for observing exotic properties of nodal line physics.

cond-mat.mtrl-sci↗

arXiv2017

Manipulating Multiple Order Parameters via Oxygen Vacancies: The case of Eu0.5Ba0.5TiO3-δ

Controlling functionalities, such as magnetism or ferroelectricity, by means of oxygen vacancies (VO) is a key issue for the future development of transition metal oxides. Progress in this field is currently addressed through VO variations and their impact on mainly one order parameter. Here we reveal a new mechanism for tuning both magnetism and ferroelectricity simultaneously by using VO. Combined experimental and density-functional theory studies of Eu0.5Ba0.5TiO3-δ, we demonstrate that oxygen vacancies create Ti3+ 3d1 defect states, mediating the ferromagnetic coupling between the localized Eu 4f7 spins, and increase an off-center displacement of Ti ions, enhancing the ferroelectric Curie temperature. The dual function of Ti sites also promises a magnetoelectric coupling in the Eu0.5Ba0.5TiO3-δ.

cond-mat.mtrl-sci↗

arXiv2017

Exploiting Web Images for Weakly Supervised Object Detection

In recent years, the performance of object detection has advanced significantly with the evolving deep convolutional neural networks. However, the state-of-the-art object detection methods still rely on accurate bounding box annotations that require extensive human labelling. Object detection without bounding box annotations, i.e, weakly supervised detection methods, are still lagging far behind. As weakly supervised detection only uses image level labels and does not require the ground truth of bounding box location and label of each object in an image, it is generally very difficult to distill knowledge of the actual appearances of objects. Inspired by curriculum learning, this paper proposes an easy-to-hard knowledge transfer scheme that incorporates easy web images to provide prior knowledge of object appearance as a good starting point. While exploiting large-scale free web imagery, we introduce a sophisticated labour free method to construct a web dataset with good diversity in object appearance. After that, semantic relevance and distribution relevance are introduced and utilized in the proposed curriculum training scheme. Our end-to-end learning with the constructed web data achieves remarkable improvement across most object classes especially for the classes that are often considered hard in other works.

cs.CV↗

arXiv2017

Prediction of triple point fermions in simple half-Heusler topological insulators

We predict the existence of triple point fermions in the band structure of several half-Heusler topological insulators by $ab~initio$ calculations and the Kane model. We find that many half-Heusler compounds exhibit multiple triple points along four independent $C_3$ axes, through which the doubly degenerate conduction bands and the nondegenerate valence band cross each other linearly nearby the Fermi energy. When projected from the bulk to the (111) surface, most of these triple points are located far away from the surface $\barΓ$ point, as distinct from previously reported triple point fermion candidates. These isolated triple points give rise to Fermi arcs on the surface, that can be readily detected by photoemission spectroscopy or scanning tunneling spectroscopy.

cond-mat.mtrl-sci↗

arXiv2017

Strong, anisotropic anomalous Hall effect and spin Hall effect in chiral antiferromagnetic compounds Mn$_3X$ ($X$ = Ge, Sn, Ga, Ir, Rh and Pt)

We have carried out a comprehensive study of the intrinsic anomalous Hall effect and spin Hall effect of several chiral antiferromagnetic compounds, Mn$_3X$ ($X$ = Ge, Sn, Ga, Ir, Rh and Pt) by $ab~initio$ band structure and Berry phase calculations. These studies reveal large and anisotropic values of both the intrinsic anomalous Hall effect and spin Hall effect. The Mn$_3X$ materials exhibit a non-collinear antiferromagnetic order which, to avoid geometrical frustration, forms planes of Mn moments that are arranged in a Kagome-type lattice. With respect to these Kagome planes, we find that both the anomalous Hall conductivity (AHC) and the spin Hall conductivity (SHC) are quite anisotropic for any of these materials. Based on our calculations, we propose how to maximize AHC and SHC for different materials. The band structures and corresponding electron filling, that we show are essential to determine the AHC and SHC, are compared for these different compounds. We point out that Mn$_3$Ga shows a large SHC of about 600 $(\hbar/e)(Ω\cdot cm)^{-1}$. Our work provides insights into the realization of strong anomalous Hall effects and spin Hall effects in chiral antiferromagetic materials.

cond-mat.mtrl-sci↗

arXiv2017

Memos: Revisiting Hybrid Memory Management in Modern Operating System

The emerging hybrid DRAM-NVM architecture is challenging the existing memory management mechanism in operating system. In this paper, we introduce memos, which can schedule memory resources over the entire memory hierarchy including cache, channels, main memory comprising DRAM and NVM simultaneously. Powered by our newly designed kernel-level monitoring module and page migration engine, memos can dynamically optimize the data placement at the memory hierarchy in terms of the on-line memory patterns, current resource utilization and feature of memory medium. Our experimental results show that memos can achieve high memory utilization, contributing to system throughput by 19.1% and QoS by 23.6% on average. Moreover, memos can reduce the NVM side memory latency by 3~83.3%, energy consumption by 25.1~99%, and benefit the NVM lifetime significantly (40X improvement on average).

cs.OS↗

arXiv2017

MIML-FCN+: Multi-instance Multi-label Learning via Fully Convolutional Networks with Privileged Information

Multi-instance multi-label (MIML) learning has many interesting applications in computer visions, including multi-object recognition and automatic image tagging. In these applications, additional information such as bounding-boxes, image captions and descriptions is often available during training phrase, which is referred as privileged information (PI). However, as existing works on learning using PI only consider instance-level PI (privileged instances), they fail to make use of bag-level PI (privileged bags) available in MIML learning. Therefore, in this paper, we propose a two-stream fully convolutional network, named MIML-FCN+, unified by a novel PI loss to solve the problem of MIML learning with privileged bags. Compared to the previous works on PI, the proposed MIML-FCN+ utilizes the readily available privileged bags, instead of hard-to-obtain privileged instances, making the system more general and practical in real world applications. As the proposed PI loss is convex and SGD compatible and the framework itself is a fully convolutional network, MIML-FCN+ can be easily integrated with state of-the-art deep learning networks. Moreover, the flexibility of convolutional layers allows us to exploit structured correlations among instances to facilitate more effective training and testing. Experimental results on three benchmark datasets demonstrate the effectiveness of the proposed MIML-FCN+, outperforming state-of-the-art methods in the application of multi-object recognition.

cs.CV↗

arXiv2016

Cloud Atlas: Discovery of Patchy Clouds and High-amplitude Rotational Modulations In a Young, Extremely Red L-type Brown Dwarf

Condensate clouds fundamentally impact the atmospheric structure and spectra of exoplanets and brown dwarfs but the connections between surface gravity, cloud structure, dust in the upper atmosphere, and the red colors of some brown dwarfs remain poorly understood. Rotational modulations enable the study of different clouds in the same atmosphere, thereby providing a method to isolate the effects of clouds. Here we present the discovery of high peak-to-peak amplitude (8%) rotational modulations in a low-gravity, extremely red (J-Ks=2.55) L6 dwarf WISEP J004701.06+680352.1 (W0047). Using the Hubble Space Telescope (HST) time-resolved grism spectroscopy we find a best-fit rotational period (13.20$\pm$0.14 hours) with a larger amplitude at 1.1 micron than at 1.7 micron. This is the third largest near-infrared variability amplitude measured in a brown dwarf, demonstrating that large-amplitude variations are not limited to the L/T transition but are present in some extremely red L-type dwarfs. We report a tentative trend between the wavelength dependence of relative amplitude, possibly proxy for small dust grains lofted in the upper atmosphere, and the likelihood of large-amplitude variability. By assuming forsterite as haze particle, we successfully explain the wavelength dependent amplitude with submicron-sized haze particles sizes of around 0.4 μm. W0047 links the earlier spectral and later spectral type brown dwarfs in which rotational modulations have been observed, the large amplitude variations in this object make this a benchmark brown dwarf for the study of cloud properties close to the L/T transition.

astro-ph.SR↗