arXiv subjects

Yue Wu

Publications and source records attributed to Yue Wu.

At least 379 records · Page 21Linked to original sources

arXiv2018

Facial Landmark Detection: a Literature Survey

The locations of the fiducial facial landmark points around facial components and facial contour capture the rigid and non-rigid facial deformations due to head movements and facial expressions. They are hence important for various facial analysis tasks. Many facial landmark detection algorithms have been developed to automatically detect those key points over the years, and in this paper, we perform an extensive review of them. We classify the facial landmark detection algorithms into three major categories: holistic methods, Constrained Local Model (CLM) methods, and the regression-based methods. They differ in the ways to utilize the facial appearance and shape information. The holistic methods explicitly build models to represent the global facial appearance and shape information. The CLMs explicitly leverage the global shape model but build the local appearance models. The regression-based methods implicitly capture facial shape and appearance information. For algorithms within each category, we discuss their underlying theories as well as their differences. We also compare their performances on both controlled and in the wild benchmark datasets, under varying facial expressions, head poses, and occlusion. Based on the evaluations, we point out their respective strengths and weaknesses. There is also a separate section to review the latest deep learning-based algorithms. The survey also includes a listing of the benchmark databases and existing software. Finally, we identify future research directions, including combining methods in different categories to leverage their respective strengths to solve landmark detection "in-the-wild".

cs.CV↗

arXiv2018

Machine Learning for Exam Triage

In this project, we extend the state-of-the-art CheXNet (Rajpurkar et al. [2017]) by making use of the additional non-image features in the dataset. Our model produced better AUROC scores than the original CheXNet.

cs.CV↗

arXiv2018

Forecasting Future Humphrey Visual Fields Using Deep Learning

Purpose: To determine if deep learning networks could be trained to forecast a future 24-2 Humphrey Visual Field (HVF). Participants: All patients who obtained a HVF 24-2 at the University of Washington. Methods: All datapoints from consecutive 24-2 HVFs from 1998 to 2018 were extracted from a University of Washington database. Ten-fold cross validation with a held out test set was used to develop the three main phases of model development: model architecture selection, dataset combination selection, and time-interval model training with transfer learning, to train a deep learning artificial neural network capable of generating a point-wise visual field prediction. Results: More than 1.7 million perimetry points were extracted to the hundredth decibel from 32,443 24-2 HVFs. The best performing model with 20 million trainable parameters, CascadeNet-5, was selected. The overall MAE for the test set was 2.47 dB (95% CI: 2.45 dB to 2.48 dB). The 100 fully trained models were able to successfully predict progressive field loss in glaucomatous eyes up to 5.5 years in the future with a correlation of 0.92 between the MD of predicted and actual future HVF (p < 2.2 x 10 -16 ) and an average difference of 0.41 dB. Conclusions: Using unfiltered real-world datasets, deep learning networks show an impressive ability to not only learn spatio-temporal HVF changes but also to generate predictions for future HVFs up to 5.5 years, given only a single HVF.

cs.CV↗

arXiv2018

Convergence Analysis and Design of Multi-block ADMM via Switched Control Theory

We consider three challenges in multi-block Alternating Direction Method of Multipliers (ADMM): building convergence conditions for ADMM with any block (variable) sequence, finding available block sequences to be fit for ADMM, and designing useful parameter controllers for ADMM with unfixed parameters. To address these challenges, we develop a switched control framework for studying multi-block ADMM. First, since ADMM recursively and alternately updates the block-variables, it is converted into a discrete-time switched dynamical system. Second, we study exponential stability and stabilizability of the switched system for linear convergence analysis and design of ADMM by employing switched Lyapunov functions. Moreover, linear matrix inequalities conditions are proposed to ensure convergence of ADMM under arbitrary sequence, to find convergent sequences, and to design the fixed parameters. These conditions are checked and solved by employing semidefinite programming. Numerical experiments further verify the effectiveness of our proposed theories.

eess.SY↗

arXiv2018

Anticipating Random Periodic Solutions--II. SPDEs with Multiplicative Linear Noise

In this paper, we study the existence of random periodic solutions for semilinear stochastic partial differential equations with multiplicative linear noise on a bounded open domain ${\cal O}\subset {\mathbb R}^d$ with smooth boundary. We identify them with the solutions of coupled forward-backward infinite horizon stochastic integral equations in $L^2({\cal O})$. We then use generalized Schauder's fixed point theorem, the relative compactness of Wiener-Sobolev spaces in $C^0([0, T], L^2(Ω\times{\cal O}))$ and a localization argument to prove the existence of solutions of the infinite horizon integral equations, which immediately implies the existence of the random periodic solution to the corresponding SPDEs. As an example, we apply our result to the stochastic Allen-Cahn equation with a periodic potential and prove the existence of a random periodic solution using a localisation argument.

math.PR↗

arXiv2018

Generating retinal flow maps from structural optical coherence tomography with artificial intelligence

Despite significant advances in artificial intelligence (AI) for computer vision, its application in medical imaging has been limited by the burden and limits of expert-generated labels. We used images from optical coherence tomography angiography (OCTA), a relatively new imaging modality that measures perfusion of the retinal vasculature, to train an AI algorithm to generate vasculature maps from standard structural optical coherence tomography (OCT) images of the same retinae, both exceeding the ability and bypassing the need for expert labeling. Deep learning was able to infer perfusion of microvasculature from structural OCT images with similar fidelity to OCTA and significantly better than expert clinicians (P < 0.00001). OCTA suffers from need of specialized hardware, laborious acquisition protocols, and motion artifacts; whereas our model works directly from standard OCT which are ubiquitous and quick to obtain, and allows unlocking of large volumes of previously collected standard OCT data both in existing clinical trials and clinical practice. This finding demonstrates a novel application of AI to medical imaging, whereby subtle regularities between different modalities are used to image the same body part and AI is used to generate detailed and accurate inferences of tissue function from structure imaging.

cs.CV↗

arXiv2018

Misclassified B stars in the Kepler field

Stellar fundamental parameters are important in the asteroseismic study of Kepler light curves. However, the most used estimates in the Kepler Input Catalog (KIC) are not accurate enough for hot stars. Using a sample of B stars from the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) spectral survey, we confirmed the systematic underestimation in KIC effective temperature and overestimation in KIC surface gravity. The good agreement between LAMOST and other follow-up observations proved the accuracy of effective temperature and surface gravity of B stars derived from LAMOST low-resolution spectra. By searching through LAMOST data, we found four misclassified main-sequence B stars in the Kepler field, which had been previously classified as A-type variables. We present spectroscopic and detailed frequency analysis of these four stars based on LAMOST spectra and Kepler photometry.

astro-ph.SR↗

arXiv2018

MRI Tumor Segmentation with Densely Connected 3D CNN

Glioma is one of the most common and aggressive types of primary brain tumors. The accurate segmentation of subcortical brain structures is crucial to the study of gliomas in that it helps the monitoring of the progression of gliomas and aids the evaluation of treatment outcomes. However, the large amount of required human labor makes it difficult to obtain the manually segmented Magnetic Resonance Imaging (MRI) data, limiting the use of precise quantitative measurements in the clinical practice. In this work, we try to address this problem by developing a 3D Convolutional Neural Network~(3D CNN) based model to automatically segment gliomas. The major difficulty of our segmentation model comes with the fact that the location, structure, and shape of gliomas vary significantly among different patients. In order to accurately classify each voxel, our model captures multi-scale contextual information by extracting features from two scales of receptive fields. To fully exploit the tumor structure, we propose a novel architecture that hierarchically segments different lesion regions of the necrotic and non-enhancing tumor~(NCR/NET), peritumoral edema~(ED) and GD-enhancing tumor~(ET). Additionally, we utilize densely connected convolutional blocks to further boost the performance. We train our model with a patch-wise training schema to mitigate the class imbalance problem. The proposed method is validated on the BraTS 2017 dataset and it achieves Dice scores of 0.72, 0.83 and 0.81 for the complete tumor, tumor core and enhancing tumor, respectively. These results are comparable to the reported state-of-the-art results, and our method is better than existing 3D-based methods in terms of compactness, time and space efficiency.

eess.IV↗

arXiv2018

Incremental Classifier Learning with Generative Adversarial Networks

In this paper, we address the incremental classifier learning problem, which suffers from catastrophic forgetting. The main reason for catastrophic forgetting is that the past data are not available during learning. Typical approaches keep some exemplars for the past classes and use distillation regularization to retain the classification capability on the past classes and balance the past and new classes. However, there are four main problems with these approaches. First, the loss function is not efficient for classification. Second, there is unbalance problem between the past and new classes. Third, the size of pre-decided exemplars is usually limited and they might not be distinguishable from unseen new classes. Forth, the exemplars may not be allowed to be kept for a long time due to privacy regulations. To address these problems, we propose (a) a new loss function to combine the cross-entropy loss and distillation loss, (b) a simple way to estimate and remove the unbalance between the old and new classes , and (c) using Generative Adversarial Networks (GANs) to generate historical data and select representative exemplars during generation. We believe that the data generated by GANs have much less privacy issues than real images because GANs do not directly copy any real image patches. We evaluate the proposed method on CIFAR-100, Flower-102, and MS-Celeb-1M-Base datasets and extensive experiments demonstrate the effectiveness of our method.

cs.CV↗

arXiv2018

The UV Emission of Stars in LAMOST Survey I. Catalogs

We present the ultraviolet magnitudes for over three million stars in the LAMOST survey, in which 2,202,116 stars are detected by $GALEX$. For 889,235 undetected stars, we develop a method to estimate their upper limit magnitudes. The distribution of (FUV $-$ NUV) shows that the color declines with increasing effective temperature for stars hotter than 7000 K in our sample, while the trend disappears for the cooler stars due to upper atmosphere emission from the regions higher than their photospheres. For stars with valid stellar parameters, we calculate the UV excesses with synthetic model spectra, and find that the (FUV $-$ NUV) vs. $R'_{\mathrm{FUV}}$ can be fitted with a linear relation and late-type dwarfs tend to have high UV excesses. There are 87,178 and 1,498,103 stars detected more than once in the visit exposures of $GALEX$ in the FUV and NUV, respectively. We make use of the quantified photometric errors to determine statistical properties of the UV variation, including intrinsic variability and the structure function on the timescale of days. The overall occurrence of possible false positives is below 1.3\% in our sample. UV absolute magnitudes are calculated for stars with valid parallaxes, which could serve as a possible reference frame in the NUV. We conclude that the colors related to UV provide good criteria to distinguish between M giants and M dwarfs, and the variability of RR Lyrae stars in our sample is stronger than that of other A and F stars.

astro-ph.SR↗

arXiv2018

A randomized and fully discrete Galerkin finite element method for semilinear stochastic evolution equations

In this paper the numerical solution of non-autonomous semilinear stochastic evolution equations driven by an additive Wiener noise is investigated. We introduce a novel fully discrete numerical approximation that combines a standard Galerkin finite element method with a randomized Runge-Kutta scheme. Convergence of the method to the mild solution is proven with respect to the $L^p$-norm, $p \in [2,\infty)$. We obtain the same temporal order of convergence as for Milstein-Galerkin finite element methods but without imposing any differentiability condition on the nonlinearity. The results are extended to also incorporate a spectral approximation of the driving Wiener process. An application to a stochastic partial differential equation is discussed and illustrated through a numerical experiment.

math.NA↗

arXiv2018

Seismic-Net: A Deep Densely Connected Neural Network to Detect Seismic Events

One of the risks of large-scale geologic carbon sequestration is the potential migration of fluids out of the storage formations. Accurate and fast detection of this fluids migration is not only important but also challenging, due to the large subsurface uncertainty and complex governing physics. Traditional leakage detection and monitoring techniques rely on geophysical observations including seismic. However, the resulting accuracy of these methods is limited because of indirect information they provide requiring expert interpretation, therefore yielding in-accurate estimates of leakage rates and locations. In this work, we develop a novel machine-learning detection package, named "Seismic-Net", which is based on the deep densely connected neural network. To validate the performance of our proposed leakage detection method, we employ our method to a natural analog site at Chimayó, New Mexico. The seismic events in the data sets are generated because of the eruptions of geysers, which is due to the leakage of $\mathrm{CO}_\mathrm{2}$. In particular, we demonstrate the efficacy of our Seismic-Net by formulating our detection problem as an event detection problem with time series data. A fixed-length window is slid throughout the time series data and we build a deep densely connected network to classify each window to determine if a geyser event is included. Through our numerical tests, we show that our model achieves precision/recall as high as 0.889/0.923. Therefore, our Seismic-Net has a great potential for detection of $\mathrm{CO}_\mathrm{2}$ leakage.

eess.SP↗

arXiv2017

Cascaded Region-based Densely Connected Network for Event Detection: A Seismic Application

Automatic event detection from time series signals has wide applications, such as abnormal event detection in video surveillance and event detection in geophysical data. Traditional detection methods detect events primarily by the use of similarity and correlation in data. Those methods can be inefficient and yield low accuracy. In recent years, because of the significantly increased computational power, machine learning techniques have revolutionized many science and engineering domains. In this study, we apply a deep-learning-based method to the detection of events from time series seismic signals. However, a direct adaptation of the similar ideas from 2D object detection to our problem faces two challenges. The first challenge is that the duration of earthquake event varies significantly; The other is that the proposals generated are temporally correlated. To address these challenges, we propose a novel cascaded region-based convolutional neural network to capture earthquake events in different sizes, while incorporating contextual information to enrich features for each individual proposal. To achieve a better generalization performance, we use densely connected blocks as the backbone of our network. Because of the fact that some positive events are not correctly annotated, we further formulate the detection problem as a learning-from-noise problem. To verify the performance of our detection methods, we employ our methods to seismic data generated from a bi-axial "earthquake machine" located at Rock Mechanics Laboratory, and we acquire labels with the help of experts. Through our numerical tests, we show that our novel detection techniques yield high accuracy. Therefore, our novel deep-learning-based detection methods can potentially be powerful tools for locating events from time series data in various applications.

cs.LG↗

arXiv2017

Ultrafast Synchronization via Local Observation

Rapid expansions of their size and frequent changes of their topology make it difficult to observe and analyze complex networks. We explore the properties of the Hankel matrix and propose an algorithm for calculating the final synchronization state that uses a local observation of a single node for a time period significantly shorter than the synchronization process. We find that synchronization can be achieved more quickly than the routine rhythm. This finding refines our understanding of the abundant ultrafast synchronization phenomena observed in nature, and it enables the efficient design of self-aligned robots.

physics.soc-ph↗

arXiv2017

Synthesis, Crystal Structure, Magnetic and Electronic Properties of the Caesium-based Transition Metal Halide Cs3Fe2Br9

The diversity of halide materials related to important solar energy systems such as CsPbX3 (X = Cl, Br, I) is explored by introducing the transition metal element Fe. In particular a new compound, Cs3Fe2Br9 (space group P6_3/mmc with a = 7.5427(8) and c = 18.5849(13) Å), has been synthesized and found to contain 0D face-sharing Fe2Br9 octahedral dimers. Unlike its isomorph, Cs3Bi2I9, it is black in color, has a low optical bandgap of 1.65 eV and exhibits antiferromagnetic behavior below TN = 13 K. Density functional theory calculations shed further light on these properties and also predict that the material should have anisotropic transport characteristics.

physics.chem-ph↗

arXiv2017

Two twin binaries with nearly identical components: KIC 4826439 and KIC 6045264

Two twin binaries, KIC 4826439 and KIC 6045264, with very similar component stars were found photometrically based on $\textit{Kepler}$ eclipsing binary light curves. The absolute parameters of the massive components are 1.156(0.03)$M_\odot$, 1.881(0.02)$R_\odot$, 6065K for KIC 4826439, and 0.874(0.3)$M_\odot$, 1.206(0.02)$R_\odot$, 6169(30)K for KIC 6045264. The differences between the components are less than two percents for all the parameters. A very low proportion of the twin binaries ($2/1592\approx0.13\%$) was found, which does not support the previous findings of the excesses of twins on binary mass ratio distribution, but support a deficiently low proportion of twins. A new method is practiced to work out the absolute parameters of the two twins without the radial velocities. This method requires the solution of the light curves, the spectra and the evolutionary isochrones of covering the complete stellar parameter space, simultaneously. We also studied their evolution tracks that: KIC 4826439 will experience an unstable mass transfer stage followed by an unclear ending, and KIC 6045264 will become a single star via an over-contact phase. It seems highly unlikely that the two twin binaries will produce twin degenerate binaries, although they have quite similar components.

astro-ph.SR↗

arXiv2017

Robust Facial Landmark Detection under Significant Head Poses and Occlusion

There have been tremendous improvements for facial landmark detection on general "in-the-wild" images. However, it is still challenging to detect the facial landmarks on images with severe occlusion and images with large head poses (e.g. profile face). In fact, the existing algorithms usually can only handle one of them. In this work, we propose a unified robust cascade regression framework that can handle both images with severe occlusion and images with large head poses. Specifically, the method iteratively predicts the landmark occlusions and the landmark locations. For occlusion estimation, instead of directly predicting the binary occlusion vectors, we introduce a supervised regression method that gradually updates the landmark visibility probabilities in each iteration to achieve robustness. In addition, we explicitly add occlusion pattern as a constraint to improve the performance of occlusion prediction. For landmark detection, we combine the landmark visibility probabilities, the local appearances, and the local shapes to iteratively update their positions. The experimental results show that the proposed method is significantly better than state-of-the-art works on images with severe occlusion and images with large head poses. It is also comparable to other methods on general "in-the-wild" images.

cs.CV↗

arXiv2017

Constrained Deep Transfer Feature Learning and its Applications

Feature learning with deep models has achieved impressive results for both data representation and classification for various vision tasks. Deep feature learning, however, typically requires a large amount of training data, which may not be feasible for some application domains. Transfer learning can be one of the approaches to alleviate this problem by transferring data from data-rich source domain to data-scarce target domain. Existing transfer learning methods typically perform one-shot transfer learning and often ignore the specific properties that the transferred data must satisfy. To address these issues, we introduce a constrained deep transfer feature learning method to perform simultaneous transfer learning and feature learning by performing transfer learning in a progressively improving feature space iteratively in order to better narrow the gap between the target domain and the source domain for effective transfer of the data from the source domain to target domain. Furthermore, we propose to exploit the target domain knowledge and incorporate such prior knowledge as a constraint during transfer learning to ensure that the transferred data satisfies certain properties of the target domain. To demonstrate the effectiveness of the proposed constrained deep transfer feature learning method, we apply it to thermal feature learning for eye detection by transferring from the visible domain. We also applied the proposed method for cross-view facial expression recognition as a second application. The experimental results demonstrate the effectiveness of the proposed method for both applications.

cs.CV↗