arXiv subjects

Song Wang

Publications and source records attributed to Song Wang.

At least 397 records · Page 22Linked to original sources

arXiv2020

Multi-Domain Multi-Task Rehearsal for Lifelong Learning

Rehearsal, seeking to remind the model by storing old knowledge in lifelong learning, is one of the most effective ways to mitigate catastrophic forgetting, i.e., biased forgetting of previous knowledge when moving to new tasks. However, the old tasks of the most previous rehearsal-based methods suffer from the unpredictable domain shift when training the new task. This is because these methods always ignore two significant factors. First, the Data Imbalance between the new task and old tasks that makes the domain of old tasks prone to shift. Second, the Task Isolation among all tasks will make the domain shift toward unpredictable directions; To address the unpredictable domain shift, in this paper, we propose Multi-Domain Multi-Task (MDMT) rehearsal to train the old tasks and new task parallelly and equally to break the isolation among tasks. Specifically, a two-level angular margin loss is proposed to encourage the intra-class/task compactness and inter-class/task discrepancy, which keeps the model from domain chaos. In addition, to further address domain shift of the old tasks, we propose an optional episodic distillation loss on the memory to anchor the knowledge for each old task. Experiments on benchmark datasets validate the proposed approach can effectively mitigate the unpredictable domain shift.

cs.LG↗

arXiv2020

Contour Transformer Network for One-shot Segmentation of Anatomical Structures

Accurate segmentation of anatomical structures is vital for medical image analysis. The state-of-the-art accuracy is typically achieved by supervised learning methods, where gathering the requisite expert-labeled image annotations in a scalable manner remains a main obstacle. Therefore, annotation-efficient methods that permit to produce accurate anatomical structure segmentation are highly desirable. In this work, we present Contour Transformer Network (CTN), a one-shot anatomy segmentation method with a naturally built-in human-in-the-loop mechanism. We formulate anatomy segmentation as a contour evolution process and model the evolution behavior by graph convolutional networks (GCNs). Training the CTN model requires only one labeled image exemplar and leverages additional unlabeled data through newly introduced loss functions that measure the global shape and appearance consistency of contours. On segmentation tasks of four different anatomies, we demonstrate that our one-shot learning method significantly outperforms non-learning-based methods and performs competitively to the state-of-the-art fully supervised deep learning methods. With minimal human-in-the-loop editing feedback, the segmentation performance can be further improved to surpass the fully supervised methods.

cs.CV↗

arXiv2020

The Statistical Characteristics of Power-Spectrum Subband Energy Ratios under Additive Gaussian White Noise

The power-spectrum subband energy ratio (PSER) has been applied in a variety of fields, but reports on its statistical properties have been limited. As such, this study investigates these characteristics in the presence of additive Gaussian white noise for both pure noise and mixed signals. By analyzing the probability and independence of power spectrum bins, and the relationship between the F and beta distributions, we develop a probability distribution for the PSER. Results showed that in the case of pure noise, the PSER follows a beta distribution. In addition, the probability density function and the quantile exhibited no relationship with the noise variance, only with the number of lines in the power spectrum, that is, PSER is not affected by noise. When Gaussian white noise was mixed with the known signal, the resulting PSER followed a doubly non-central beta distribution. In this case, it was difficult to identify the quantile, as the probability density and cumulative distribution functions were represented by infinite series. However, when a spectral bin did not contain the power spectrum of the known signal, an approximated quantile was found. This quantile is strictly proved to be in agreement with the quantile in the case of pure noise and offers a convenient methodology for identifying valid signals.

eess.SP↗

arXiv2020

Deep Poisoning: Towards Robust Image Data Sharing against Visual Disclosure

Due to respectively limited training data, different entities addressing the same vision task based on certain sensitive images may not train a robust deep network. This paper introduces a new vision task where various entities share task-specific image data to enlarge each other's training data volume without visually disclosing sensitive contents (e.g. illegal images). Then, we present a new structure-based training regime to enable different entities learn task-specific and reconstruction-proof image representations for image data sharing. Specifically, each entity learns a private Deep Poisoning Module (DPM) and insert it to a pre-trained deep network, which is designed to perform the specific vision task. The DPM deliberately poisons convolutional image features to prevent image reconstructions, while ensuring that the altered image data is functionally equivalent to the non-poisoned data for the specific vision task. Given this equivalence, the poisoned features shared from one entity could be used by another entity for further model refinement. Experimental results on image classification prove the efficacy of the proposed method.

cs.CV↗

arXiv2020

TTPLA: An Aerial-Image Dataset for Detection and Segmentation of Transmission Towers and Power Lines

Accurate detection and segmentation of transmission towers~(TTs) and power lines~(PLs) from aerial images plays a key role in protecting power-grid security and low-altitude UAV safety. Meanwhile, aerial images of TTs and PLs pose a number of new challenges to the computer vision researchers who work on object detection and segmentation -- PLs are long and thin, and may show similar color as the background; TTs can be of various shapes and most likely made up of line structures of various sparsity; The background scene, lighting, and object sizes can vary significantly from one image to another. In this paper we collect and release a new TT/PL Aerial-image (TTPLA) dataset, consisting of 1,100 images with the resolution of 3,840$\times$2,160 pixels, as well as manually labeled 8,987 instances of TTs and PLs. We develop novel policies for collecting, annotating, and labeling the images in TTPLA. Different from other relevant datasets, TTPLA supports evaluation of instance segmentation, besides detection and semantic segmentation. To build a baseline for detection and segmentation tasks on TTPLA, we report the performance of several state-of-the-art deep learning models on our dataset. TTPLA dataset is publicly available at https://github.com/r3ab/ttpla_dataset

cs.CV↗

arXiv2020

Comparisons of Different Fitting Methods for the Physical Parameters of A Star Cluster Sample of M33 with Spectroscopy and Photometry

Star clusters are good tracers for formation and evolution of galaxies. We compared different fitting methods by using spectra (or by combining photometry) to determine the physical parameters. We choose a sample of 17 star clusters in M33, which previously lacked spectroscopic observations. The low-resolution spectra were taken with the Xinglong 2.16-m reflector of NAOC. The photometry used in the fitting includes $\rm u_{SC}$ and $\rm v_{SAGE}$ bands from the SAGE survey, as well as the published $UBVRI$ and $ugriz$ photometry. We firstly derived ages and metallicities with the {\sc ULySS} (Vazdekis et al. and {\sc pegase-hr}) SSP model and the Bruzual \& Charlot (2003) (BC03) stellar population synthesis models for the full-spectrum fitting. The fitting results of both the BC03 and {\sc ULySS} models seem consistent with those of previous works as well. Then we add the SAGE $\rm u_{SC}$ and $\rm v_{SAGE}$ photometry in the spectroscopic fitting with the BC03 models. It seems the results become much better, especially for the Padova 2000+Chabrier IMF set. Finally we add more photometry data, $UBVRI$ and $ugriz$, in the fitting and we found that the results do not improve significantly. Therefore, we conclude that the photometry is useful for improving the fitting results, especially for the blue bands ($λ<4000$ Å), e.g., $\rm u_{SC}$ and $\rm v_{SAGE}$ band. At last, we discuss the "UV-excess" for the star clusters and we find five star clusters have UV-excess, based on the $GALEX$ FUV, NUV photometry.

astro-ph.GA↗

arXiv2020

Stellar X-ray activity across the Hertzsprung-Russell diagram. I. Catalogs

Stellar magnetic activity provides substantial information on the magnetic dynamo and the coronal heating process. We present a catalog of X-ray activity for about 6000 stars, based on the $Chandra$ and $Gaia$ DR2 data. We also classified more than 3000 stars as young stellar objects, dwarf stars, or giant stars. By using the stars with valid stellar parameters and classifications, we studied the distribution of X-ray luminosity ($L_X$) and the ratio of X-ray-to-bolometric luminosities ($R_X$), the positive relation between $L_X$, $R_X$, and hardness ratio, and the long-term X-ray variation. This catalog can be used to investigate some important scientific topics, including the activity-rotation relation, the comparison between different activity indicators, and the activities of interesting objects (e.g., A-type stars and giants). As an example, we use the catalog to study the activity-rotation relation, and find that the young stellar objects, dwarfs, and giants fall on a single sequence in the relation $R_X$ versus Rossby number, while the giants do not follow the relation $R_X$ versus $P_{\rm rot}^{-2}R^{-4}$ valid for dwarfs.

astro-ph.SR↗

arXiv2020

Zero-Dimensional Organic-Inorganic Hybrid Material with Ultra-Narrow-Red Emission at Room Temperature

Recently, low-dimensional organic-inorganic hybrid halide compounds have aroused great attention in the optoelectronic field, due to the unique topology and optical properties. Herein, we report an Mn4+ doped [N(CH3)4]2TiF6 zero-dimensional organic-inorganic hybrid phosphor, which could not only exhibit very narrow and pure red emission, but also maintain efficient emission intensity at room temperature. The crystal structure, photoluminescence properties and temperature sensing application are discussed. The excellent temperature dependent luminescent properties are attributed to the rigid structure and isolated MnF62- octahedra in the total crystal framework. These results will help design suitable materials and devices in both warm white light emitting diodes and optical sensors.

cond-mat.mtrl-sci↗

arXiv2020

Learning to Segment Anatomical Structures Accurately from One Exemplar

Accurate segmentation of critical anatomical structures is at the core of medical image analysis. The main bottleneck lies in gathering the requisite expert-labeled image annotations in a scalable manner. Methods that permit to produce accurate anatomical structure segmentation without using a large amount of fully annotated training images are highly desirable. In this work, we propose a novel contribution of Contour Transformer Network (CTN), a one-shot anatomy segmentor including a naturally built-in human-in-the-loop mechanism. Segmentation is formulated by learning a contour evolution behavior process based on graph convolutional networks (GCNs). Training of our CTN model requires only one labeled image exemplar and leverages additional unlabeled data through newly introduced loss functions that measure the global shape and appearance consistency of contours. We demonstrate that our one-shot learning method significantly outperforms non-learning-based methods and performs competitively to the state-of-the-art fully supervised deep learning approaches. With minimal human-in-the-loop editing feedback, the segmentation performance can be further improved and tailored towards the observer desired outcomes. This can facilitate the clinician designed imaging-based biomarker assessments (to support personalized quantitative clinical diagnosis) and outperforms fully supervised baselines.

cs.CV↗

arXiv2020

Shadow Removal by a Lightness-Guided Network with Training on Unpaired Data

Shadow removal can significantly improve the image visual quality and has many applications in computer vision. Deep learning methods based on CNNs have become the most effective approach for shadow removal by training on either paired data, where both the shadow and underlying shadow-free versions of an image are known, or unpaired data, where shadow and shadow-free training images are totally different with no correspondence. In practice, CNN training on unpaired data is more preferred given the easiness of training data collection. In this paper, we present a new Lightness-Guided Shadow Removal Network (LG-ShadowNet) for shadow removal by training on unpaired data. In this method, we first train a CNN module to compensate for the lightness and then train a second CNN module with the guidance of lightness information from the first CNN module for final shadow removal. We also introduce a loss function to further utilise the colour prior of existing data. Extensive experiments on widely used ISTD, adjusted ISTD and USR datasets demonstrate that the proposed method outperforms the state-of-the-art methods with training on unpaired data.

cs.CV↗

arXiv2020

A Catalog of Short Period Spectroscopic and Eclipsing Binaries Identified from the LAMOST & PTF Surveys

Binaries play key roles in determining stellar parameters and exploring stellar evolution models. We build a catalog of 88 eclipsing binaries with spectroscopic information, taking advantage of observations from both the Large Sky Area Multi-Object fiber Spectroscopic Telescope (LAMOST) and the Palomar Transient Factory (PTF) surveys. A software pipeline is constructed to identify binary candidates by examining their light curves. The orbital periods of binaries are derived from the Lomb-Scargle method. The key distinguishing features of eclipsing binaries are recognized by a new filter \textit{Flat Test}. We classify the eclipsing binaries by applying Fourier analysis on the light curves. Among all the binary stars, 13 binaries are identified as eclipsing binaries for the first time. The catalog contains information: position, primary eclipsing magnitude and time, eclipsing depth, the number of photometry and radial velocity observations, largest radial velocity difference, binary type, the effective temperature of observable star $T_{\rm eff}$, and surface gravity of observable star log \emph{g}. The false-positive probability is calculated by using both a Monte Carlo simulation and real data from the SDSS Stripe 82 Standard Catalog. The binaries in the catalog are mostly with a period of less than one day. The period distribution shows a 0.22-day cut-off which is consistent with the low probability of an eclipsing binary rotating with such a period.

astro-ph.SR↗

arXiv2020

Phase-dependent study of near-infrared disk emission lines in LB-1

The mass, origin and evolutionary stage of the binary system LB-1 has been the subject of intense debate, following the claim that it hosts an $\sim$70$M_{\odot}$ black hole, in stark contrast with the expectations for stellar remnants in the Milky Way. We conducted a high-resolution, phase-resolved spectroscopic study of the near-infrared Paschen lines in this system, using the 3.5-m telescope at Calar Alto Observatory. We find that Pa$β$ and Pa$γ$ (after proper subtraction of the stellar absorption component) are well fitted with a standard double-peaked model, typical of disk emission. We measured the velocity shifts of the red and blue peaks at 28 orbital phases: the line center has an orbital motion in perfect antiphase with the stellar motion, and the radial velocity amplitude ranges from 8 to 13 km/s for different choices of lines and profile modelling. We interpret this curve as proof that the disk is tracing the orbital motion of the primary, ruling out the circumbinary disk and the hierarchical triple scenarios. The phase-averaged peak-to-peak half-separation (proxy for the projected rotational velocity of the outer disk) is $\sim$70 km s$^{-1}$, larger than the stellar orbital velocity and also inconsistent with a circumbinary disk. From those results, we infer a primary mass 4--8 times higher than the secondary mass. Moreover, we show that the ratio of the blue and red peaks (V/R intensity ratio) has a sinusoidal behaviour in phase with the secondary star, which can be interpreted as the effect of external irradiation by the secondary star on the outer disk. Finally, we briefly discuss our findings in the context of alternative scenarios recently proposed for LB-1. Definitive tests between alternative solutions will require further astrometric data from $Gaia$.

astro-ph.SR↗

arXiv2020

Modeling Cross-view Interaction Consistency for Paired Egocentric Interaction Recognition

With the development of Augmented Reality (AR), egocentric action recognition (EAR) plays important role in accurately understanding demands from the user. However, EAR is designed to help recognize human-machine interaction in single egocentric view, thus difficult to capture interactions between two face-to-face AR users. Paired egocentric interaction recognition (PEIR) is the task to collaboratively recognize the interactions between two persons with the videos in their corresponding views. Unfortunately, existing PEIR methods always directly use linear decision function to fuse the features extracted from two corresponding egocentric videos, which ignore consistency of interaction in paired egocentric videos. The consistency of interactions in paired videos, and features extracted from them are correlated to each other. On top of that, we propose to build the relevance between two views using biliear pooling, which capture the consistency of two views in feature-level. Specifically, each neuron in the feature maps from one view connects to the neurons from another view, which guarantee the compact consistency between two views. Then all possible paired neurons are used for PEIR for the inside consistent information of them. To be efficient, we use compact bilinear pooling with Count Sketch to avoid directly computing outer product in bilinear. Experimental results on dataset PEV shows the superiority of the proposed methods on the task PEIR.

cs.CV↗

arXiv2020

MUTATT: Visual-Textual Mutual Guidance for Referring Expression Comprehension

Referring expression comprehension (REC) aims to localize a text-related region in a given image by a referring expression in natural language. Existing methods focus on how to build convincing visual and language representations independently, which may significantly isolate visual and language information. In this paper, we argue that for REC the referring expression and the target region are semantically correlated and subject, location and relationship consistency exist between vision and language.On top of this, we propose a novel approach called MutAtt to construct mutual guidance between vision and language, which treat vision and language equally thus yield compact information matching. Specifically, for each module of subject, location and relationship, MutAtt builds two kinds of attention-based mutual guidance strategies. One strategy is to generate vision-guided language embedding for the sake of matching relevant visual feature. The other reversely generates language-guided visual feature to match relevant language embedding. This mutual guidance strategy can effectively guarantees the vision-language consistency in three modules. Experiments on three popular REC datasets demonstrate that the proposed approach outperforms the current state-of-the-art methods.

cs.CV↗

arXiv2020

Label-guided Learning for Text Classification

Text classification is one of the most important and fundamental tasks in natural language processing. Performance of this task mainly dependents on text representation learning. Currently, most existing learning frameworks mainly focus on encoding local contextual information between words. These methods always neglect to exploit global clues, such as label information, for encoding text information. In this study, we propose a label-guided learning framework LguidedLearn for text representation and classification. Our method is novel but simple that we only insert a label-guided encoding layer into the commonly used text representation learning schemas. That label-guided layer performs label-based attentive encoding to map the universal text embedding (encoded by a contextual information learner) into different label spaces, resulting in label-wise embeddings. In our proposed framework, the label-guided layer can be easily and directly applied with a contextual encoding method to perform jointly learning. Text information is encoded based on both the local contextual information and the global label clues. Therefore, the obtained text embeddings are more robust and discriminative for text classification. Extensive experiments are conducted on benchmark datasets to illustrate the effectiveness of our proposed method.

cs.CL↗

arXiv2019

Machine Learning Regression of extinction in the second $Gaia$ Data Release

Machine learning has become a popular tool to help us make better decisions and predictions, based on experiences, observations and analysing patterns within a given data set without explicitly functions. In this paper, we describe an application of the supervised machine-learning algorithm to the extinction regression for the second Gaia data release, based on the combination of Large Sky Area Multi-Object Fiber Spectroscopic Telescope, Sloan Extension for Galactic Understanding and Exploration and the Apache Point Observatory Galactic Evolution Experiment. The derived extinction in our training sample is consistent with other spectrum-based estimates, and its standard deviation of the cross validations is 0.0127 mag. A blind test is carried out using the RAdial Velocity Experiment catalog, and the standard deviation is 0.0372 mag. Such precise training sample enable us to regress the extinction, E(BP-RP), for 133 million stars in the second Gaia data release. Of these, 106 million stars have the uncertainties less than 0.1 mag, which suffer less bias from the external regression. We also find that there are high deviations between the extinctions form photometry-based methods, and between spectrum- and photometry-based methods. This implies that spectrum-based method could bring more signal to a regressing model than multi-band photometry, and a higher signal-to-noise ratio would acquire a more reliable result.

astro-ph.SR↗

arXiv2019

A wide star-black-hole binary system from radial-velocity measurements

All stellar mass black holes have hitherto been identified by X-rays emitted by gas that is accreting onto the black hole from a companion star. These systems are all binaries with black holes below 30 M$_{\odot}$$^{1-4}$. Theory predicts, however, that X-ray emitting systems form a minority of the total population of star-black hole binaries$^{5,6}$. When the black hole is not accreting gas, it can be found through radial velocity measurements of the motion of the companion star. Here we report radial velocity measurements of a Galactic star, LB-1, which is a B-type star, taken over two years. We find that the motion of the B-star and an accompanying H$α$ emission line require the presence of a dark companion with a mass of $68^{+11}_{-13}$ M$_{\odot}$, which can only be a black hole. The long orbital period of 78.9 days shows that this is a wide binary system. The gravitational wave experiments have detected similarly massive black holes$^{7,8}$, but forming such massive ones in a high-metallicity environment would be extremely challenging to current stellar evolution theories$^{9-11}$.

astro-ph.SR↗

arXiv2019

Improved Deep Hashing with Soft Pairwise Similarity for Multi-label Image Retrieval

Hash coding has been widely used in the approximate nearest neighbor search for large-scale image retrieval. Recently, many deep hashing methods have been proposed and shown largely improved performance over traditional feature-learning-based methods. Most of these methods examine the pairwise similarity on the semantic-level labels, where the pairwise similarity is generally defined in a hard-assignment way. That is, the pairwise similarity is '1' if they share no less than one class label and '0' if they do not share any. However, such similarity definition cannot reflect the similarity ranking for pairwise images that hold multiple labels. In this paper, a new deep hashing method is proposed for multi-label image retrieval by re-defining the pairwise similarity into an instance similarity, where the instance similarity is quantified into a percentage based on the normalized semantic labels. Based on the instance similarity, a weighted cross-entropy loss and a minimum mean square error loss are tailored for loss-function construction, and are efficiently used for simultaneous feature learning and hash coding. Experiments on three popular datasets demonstrate that, the proposed method outperforms the competing methods and achieves the state-of-the-art performance in multi-label image retrieval.

cs.CV↗