Searcharxiv⌕ Search

arXiv subjects

Bo Wang

Publications and source records attributed to Bo Wang.

At least 703 records · Page 39Linked to original sources
arXiv2018

Network Enhancement: a general method to denoise weighted biological networks

Networks are ubiquitous in biology where they encode connectivity patterns at all scales of organization, from molecular to the biome. However, biological networks are noisy due to the limitations of measurement technology and inherent natural variation, which can hamper discovery of network patterns and dynamics. We propose Network Enhancement (NE), a method for improving the signal-to-noise ratio of undirected, weighted networks. NE uses a doubly stochastic matrix operator that induces sparsity and provides a closed-form solution that increases spectral eigengap of the input network. As a result, NE removes weak edges, enhances real connections, and leads to better downstream performance. Experiments show that NE improves gene function prediction by denoising tissue-specific interaction networks, alleviates interpretation of noisy Hi-C contact maps from the human genome, and boosts fine-grained identification accuracy of species. Our results indicate that NE is widely applicable for denoising biological networks.

q-bio.MN↗
arXiv2018

Information Encoding with Optical Dielectric Metasurface via Independent Multichannels

Information encryption and security is a prerequisite for information technology which can be realized by optical metasurface owing to its arbitrary manipulation over the wavelength, polarization, phase and amplitude of light. So far information encoding can be implemented by the metasurface in one dimensional (1D) mode (either wavelength or polarization) only with several combinations of independent channels. Here we successfully apply dielectric metasurfaces in a 2D mode (both wavelength and polarization) with far more combinations of independent channels to encrypt information, which therefore enhances the encryption security dramatically. Six independent channels by two circular polarization states (RCP and LCP) and three visible wavelengths (633 nm, 532 nm and 473 nm) in 2D mode can produce 63 combinations available to information encoding, in sharp contrast with 7 combinations by 3 independent channels in 1D mode. This 2D mode encoding strategy paves a novel pathway for escalating the security level of information in multichannel information encryption, anti-counterfeiting, optical data storage, and information processing.

physics.optics↗
arXiv2018

Movie Question Answering: Remembering the Textual Cues for Layered Visual Contents

Movies provide us with a mass of visual content as well as attracting stories. Existing methods have illustrated that understanding movie stories through only visual content is still a hard problem. In this paper, for answering questions about movies, we put forward a Layered Memory Network (LMN) that represents frame-level and clip-level movie content by the Static Word Memory module and the Dynamic Subtitle Memory module, respectively. Particularly, we firstly extract words and sentences from the training movie subtitles. Then the hierarchically formed movie representations, which are learned from LMN, not only encode the correspondence between words and visual content inside frames, but also encode the temporal alignment between sentences and frames inside movie clips. We also extend our LMN model into three variant frameworks to illustrate the good extendable capabilities. We conduct extensive experiments on the MovieQA dataset. With only visual content as inputs, LMN with frame-level representation obtains a large performance improvement. When incorporating subtitles into LMN to form the clip-level representation, we achieve the state-of-the-art performance on the online evaluation task of 'Video+Subtitles'. The good performance successfully demonstrates that the proposed framework of LMN is effective and the hierarchically formed movie representations have good potential for the applications of movie question answering.

cs.CV↗
arXiv2018

Single-Shot Object Detection with Enriched Semantics

We propose a novel single shot object detection network named Detection with Enriched Semantics (DES). Our motivation is to enrich the semantics of object detection features within a typical deep detector, by a semantic segmentation branch and a global activation module. The segmentation branch is supervised by weak segmentation ground-truth, i.e., no extra annotation is required. In conjunction with that, we employ a global activation module which learns relationship between channels and object classes in a self-supervised manner. Comprehensive experimental results on both PASCAL VOC and MS COCO detection datasets demonstrate the effectiveness of the proposed method. In particular, with a VGG16 based DES, we achieve an mAP of 81.7 on VOC2007 test and an mAP of 32.8 on COCO test-dev with an inference speed of 31.5 milliseconds per image on a Titan Xp GPU. With a lower resolution version, we achieve an mAP of 79.7 on VOC2007 with an inference speed of 13.0 milliseconds per image.

cs.CV↗
arXiv2018

Political Homophily in Independence Movements: Analysing and Classifying Social Media Users by National Identity

Social media and data mining are increasingly being used to analyse political and societal issues. Here we undertake the classification of social media users as supporting or opposing ongoing independence movements in their territories. Independence movements occur in territories whose citizens have conflicting national identities; users with opposing national identities will then support or oppose the sense of being part of an independent nation that differs from the officially recognised country. We describe a methodology that relies on users' self-reported location to build large-scale datasets for three territories -- Catalonia, the Basque Country and Scotland. An analysis of these datasets shows that homophily plays an important role in determining who people connect with, as users predominantly choose to follow and interact with others from the same national identity. We show that a classifier relying on users' follow networks can achieve accurate, language-independent classification performances ranging from 85% to 97% for the three territories.

cs.CL↗
arXiv2018

Deep Co-Training for Semi-Supervised Image Recognition

In this paper, we study the problem of semi-supervised image recognition, which is to learn classifiers using both labeled and unlabeled images. We present Deep Co-Training, a deep learning based method inspired by the Co-Training framework. The original Co-Training learns two classifiers on two views which are data from different sources that describe the same instances. To extend this concept to deep learning, Deep Co-Training trains multiple deep neural networks to be the different views and exploits adversarial examples to encourage view difference, in order to prevent the networks from collapsing into each other. As a result, the co-trained networks provide different and complementary information about the data, which is necessary for the Co-Training framework to achieve good results. We test our method on SVHN, CIFAR-10/100 and ImageNet datasets, and our method outperforms the previous state-of-the-art methods by a large margin.

cs.CV↗
arXiv2018

The progenitors of type Ia supernovae in the semidetached binaries with red giant donors

Context. The companions of the exploding carbon-oxygen white dwarfs (CO WDs) for producing type Ia supernovae (SNe Ia) are still not conclusively confirmed. A red-giant (RG) star has been suggested to be the mass donor of the exploding WD, named as the symbiotic channel. However, previous studies on the this channel gave a relatively low rate of SNe Ia. Aims. We aim to systematically investigate the parameter space, Galactic rates and delay time distributions of SNe Ia from the symbiotic channel by employing a revised mass-transfer prescription. Methods. We adopted an integrated mass-transfer prescription to calculate the mass-transfer process from a RG star onto the WD. In this prescription, the mass-transfer rate varies with the local material states. Results. We evolved a large number of WD+RG systems, and found that the parameter space of WD+RG systems for producing SNe Ia is significantly enlarged. This channel could produce SNe Ia with intermediate and old ages, contributing to at most 5% of all SNe Ia in the Galaxy. Our model increases the SN Ia rate from this channel by a factor of 5. We suggest that the symbiotic systems RS Oph and T CrB are strong candidates for the progenitors of SNe Ia.

astro-ph.SR↗
arXiv2018

Evolving ONe WD+He star systems to intermediate-mass binary pulsars

It has been suggested that accretion-induced collapse (AIC) is a non-negligible path for the formation of the observed neutron stars (NSs). An ONe white dwarf (WD) that accretes material from a He star may experience AIC process and eventually produce intermediate-mass binary pulsars (IMBPs), named as the ONe WD+He star scenario. Note that previous studies can only account for part of the observed IMBPs with short orbital periods. In this work, we investigate the evolution of about 900 ONe WD+He star binaries to explore the distribution of IMBPs. We found that the ONe WD+He star scenario could form IMBPs including pulsars with 5-340 ms spin periods and 0.75-1.38 Msun WD companions, in which the orbital periods range from 0.04 to 900 d. Compared with the 20 observed IMBPs, this scenario can cover the parameters of 13 sources in the final orbital period-WD mass plane and the Corbet diagram, most of which has short orbital periods. We found that the ONe WD+He star scenario can explain almost all the observed IMBPs with short orbital periods. This work can well match the observed parameters of PSR J1802-2124 (one of the two precisely observed IMBPs), providing a possible evolutional path for its formation. We also speculate that the compact companion of HD 49798 (a hydrogen depleted sdO6 star) may be not a NS based on the present work.

astro-ph.SR↗
arXiv2018

Learning to Synthesize

In many scenarios we need to find the most likely program under a local context, where the local context can be an incomplete program, a partial specification, natural language description, etc. We call such problem program estimation. In this paper we propose an abstract framework, learning to synthesis, or L2S in short, to address this problem. L2S combines four tools to achieve this: syntax is used to define the search space and search steps, constraints are used to prune off invalid candidates at each search step, machine-learned models are used to estimate conditional probabilities for the candidates at each search step, and search algorithms are used to find the best possible solution. The main goal of L2S is to lay out the design space to motivate the research on program estimation. We have performed a preliminary evaluation by instantiating this framework for synthesizing conditions of an automated program repair (APR) system. The training data are from the project itself and related JDK packages. Compared to ACS, a state-of-the-art condition synthesis system for program repair, our approach could deal with a larger search space such that we fixed 4 additional bugs outside the search space of ACS, and relies only on the source code of the current projects.

cs.SE↗
arXiv2018

Mass-accreting white dwarfs and type Ia supernovae

Type Ia supernovae (SNe Ia) play a prominent role in understanding the evolution of Universe. They are thought to be thermonuclear explosions of mass-accreting carbon-oxygen white dwarfs (CO WDs) in binaries, although the mass donors of the accreting WDs are still not well determined. In this article, I review recent studies on mass-accreting WDs, including H- and He-accreting WDs. I also review currently most studied progenitor models of SNe Ia, i.e., the single-degenerate model (including the WD+MS channel, the WD+RG channel and the WD+He star channel), the double-degenerate model (including the violent merger scenario) and the sub-Chandrasekhar mass model. Recent progress on these progenitor models is discussed, including the initial parameter space for producing SNe Ia, the binary evolutionary paths to SNe Ia, the progenitor candidates of SNe Ia, the possible surviving companion stars of SNe Ia, and some observational constraints, etc. Some other potential progenitor models of SNe Ia are also summarized, including the hybrid CONe WD model, the core-degenerate model, the double WD collision model, the spin-up/spin-down model, and the model of WDs near black holes.To date, it seems that two or more progenitor models are needed to explain the observed diversity among SNe Ia.

astro-ph.SR↗
arXiv2018

Gradually Updated Neural Networks for Large-Scale Image Recognition

Depth is one of the keys that make neural networks succeed in the task of large-scale image recognition. The state-of-the-art network architectures usually increase the depths by cascading convolutional layers or building blocks. In this paper, we present an alternative method to increase the depth. Our method is by introducing computation orderings to the channels within convolutional layers or blocks, based on which we gradually compute the outputs in a channel-wise manner. The added orderings not only increase the depths and the learning capacities of the networks without any additional computation costs, but also eliminate the overlap singularities so that the networks are able to converge faster and perform better. Experiments show that the networks based on our method achieve the state-of-the-art performances on CIFAR and ImageNet datasets.

cs.CV↗
arXiv2018

Accreting CO material onto ONe white dwarfs towards accretion-induced collapse

The final outcomes of accreting ONe white dwarfs (ONe WDs) have been studied for several decades, but there are still some issues not resolved. Recently, some studies suggested that the deflagration of oxygen would occur for accreting ONe WDs with Chandrasekhar masses. In this paper, we aim to investigate whether ONe WDs can experience accretion-induced collapse (AIC) or explosions when their masses approach the Chandrasekhar limit. Employing the stellar evolution code modules for experiments in stellar astrophysics (MESA), we simulate the long-term evolution of ONe WDs by accreting CO material. The ONe WDs undergo weak multicycle carbon flashes during the mass-accretion process, leading to the mass increase of the WDs. We found that different initial WD masses and mass-accretion rates have influence on the evolution of central density and temperature. However, the central temperature cannot reach the explosive oxygen ignition temperature due to the neutrino cooling. This work implies that the final outcome of accreting ONe WDs is electron-capture induced collapse rather than thermonuclear explosion.

astro-ph.SR↗
arXiv2018

Carbon stars identified from LAMOST DR4 using Machine Learning

In this work, we present a catalog of 2651 carbon stars from the fourth Data Release (DR4) of the Large Sky Area Multi-Object Fiber Spectroscopy Telescope (LAMOST). Using an efficient machine-learning algorithm, we find out these stars from more than seven million spectra. As a by-product, 17 carbon-enhanced metal-poor (CEMP) turnoff star candidates are also reported in this paper, and they are preliminarily identified by their atmospheric parameters. Except for 176 stars that could not be given spectral types, we classify the other 2475 carbon stars into five subtypes including 864 C-H, 226 C-R, 400 C-J, 266 C-N, and 719 barium stars based on a series of spectral features. Furthermore, we divide the C-J stars into three subtypes of CJ( H), C-J(R), C-J(N), and about 90% of them are cool N-type stars as expected from previous literature. Beside spectroscopic classification, we also match these carbon stars to multiple broadband photometries. Using ultraviolet photometry data, we find that 25 carbon stars have FUV detections and they are likely to be in binary systems with compact white dwarf companions.

astro-ph.SR↗
arXiv2018

SIMLR: A Tool for Large-Scale Genomic Analyses by Multi-Kernel Learning

We here present SIMLR (Single-cell Interpretation via Multi-kernel LeaRning), an open-source tool that implements a novel framework to learn a sample-to-sample similarity measure from expression data observed for heterogenous samples. SIMLR can be effectively used to perform tasks such as dimension reduction, clustering, and visualization of heterogeneous populations of samples. SIMLR was benchmarked against state-of-the-art methods for these three tasks on several public datasets, showing it to be scalable and capable of greatly improving clustering performance, as well as providing valuable insights by making the data more interpretable via better a visualization. Availability and Implementation SIMLR is available on GitHub in both R and MATLAB implementations. Furthermore, it is also available as an R package on http://bioconductor.org.

q-bio.GN↗
arXiv2017

Ultrafast coherent energy transfer with high efficiency based on plasmonic nanostructures

The theory of energy transfer dynamics of a pair of donor and acceptor molecules located in the plasmonic hot spots is developed by means of the master equation approach and the electromagnetic Green's tensor technique. A nonlocal effect has been considered by using a hydrodynamic model. The coherent interaction between the two molecules in plasmonic nanostructures is investigated under some strong coupling conditions. It is shown that the energy transfer efficiency of a pair of molecules can be improved largely and the transfer time decreases to dozens of femtoseconds when the contribution of quantum coherence is considered. The physical origin for such a phenomenon has also been analyzed. This ultrafast and high-efficiency energy transfer mechanism could be beneficial for artificial light-harvesting devices.

physics.optics↗
arXiv2017

Observing half and integer polarization vortices at band degeneracies

Far-field polarization vortices were recently found on singlet bands in the momentum-space of two-dimensional photonic lattices, also known as the dark states and bound states in continuum. Here, we theoretically proposed and experimentally verified the existence of the polarization vortices at the degenerate points of photonic dispersions, whose vortex cores can be radiative bright states. Half-charged vortices were generated from the Dirac points of π Berry phase and integer-charged vortices were generated from a quadratic degeneracy. Using a home-made polarization-resolved momentum-space imaging spectroscopy, we observed the complete evolution of the splitting from one quadratic point to a pair of Dirac cones by tracking the winding of the polarization vectors and the full spectrum of iso-frequency contours.

physics.optics↗
arXiv2017

Deep Regression Forests for Age Estimation

Age estimation from facial images is typically cast as a nonlinear regression problem. The main challenge of this problem is the facial feature space w.r.t. ages is heterogeneous, due to the large variation in facial appearance across different persons of the same age and the non-stationary property of aging patterns. In this paper, we propose Deep Regression Forests (DRFs), an end-to-end model, for age estimation. DRFs connect the split nodes to a fully connected layer of a convolutional neural network (CNN) and deal with heterogeneous data by jointly learning input-dependant data partitions at the split nodes and data abstractions at the leaf nodes. This joint learning follows an alternating strategy: First, by fixing the leaf nodes, the split nodes as well as the CNN parameters are optimized by Back-propagation; Then, by fixing the split nodes, the leaf nodes are optimized by iterating a step-size free and fast-converging update rule derived from Variational Bounding. We verify the proposed DRFs on three standard age estimation benchmarks and achieve state-of-the-art results on all of them.

cs.CV↗