arXiv subjects

Xiang Li

Publications and source records attributed to Xiang Li.

At least 1,207 records · Page 67Linked to original sources

arXiv2021

PC-GAIN: Pseudo-label Conditional Generative Adversarial Imputation Networks for Incomplete Data

Datasets with missing values are very common in real world applications. GAIN, a recently proposed deep generative model for missing data imputation, has been proved to outperform many state-of-the-art methods. But GAIN only uses a reconstruction loss in the generator to minimize the imputation error of the non-missing part, ignoring the potential category information which can reflect the relationship between samples. In this paper, we propose a novel unsupervised missing data imputation method named PC-GAIN, which utilizes potential category information to further enhance the imputation power. Specifically, we first propose a pre-training procedure to learn potential category information contained in a subset of low-missing-rate data. Then an auxiliary classifier is determined using the synthetic pseudo-labels. Further, this classifier is incorporated into the generative adversarial framework to help the generator to yield higher quality imputation results. The proposed method can improve the imputation quality of GAIN significantly. Experimental results on various benchmark datasets show that our method is also superior to other baseline approaches. Our code is available at \url{https://github.com/WYu-Feng/pc-gain}.

cs.LG↗

arXiv2021

Enhancing interferometer sensitivity without sacrificing bandwidth and stability: beyond single-mode and resolved-sideband approximation

Quantum noise limits the sensitivity of precision measurement devices, such as laser interferometer gravitational-wave observatories and axion detectors. In the shot-noise-limited regime, these resonant detectors are subject to a trade-off between the peak sensitivity and bandwidth. One approach to circumvent this limitation in gravitational-wave detectors is to embed an anomalous-dispersion optomechanical filter to broaden the bandwidth. The original filter cavity design, however, makes the entire system unstable. Recently, we proposed the coherent feedback between the arm cavity and the optomechanical filter to eliminate the instability via PT-symmetry. The original analysis based upon the Hamiltonian formalism adopted the single-mode and resolved-sideband approximations. In this paper, we go beyond these approximations and consider realistic parameters. We show that the main conclusion concerning stability remains intact, with both Nyquist analysis and a detailed time-domain simulation.

quant-ph↗

arXiv2021

Black Hole Mass Function of Coalescing Binary Black Hole Systems: Is There a Pulsational Pair Instability Mass Cutoff?

We analyze the LIGO/Virgo GWTC-2 catalog to study the primary mass distribution of the merging black holes. We perform hierarchical Bayesian analysis, and examine whether the mass distribution has a sharp cutoff for primary black hole masses below $65 M_\odot$, as predicted in pulsational pair instability supernova model. We construct two empirical mass functions. One is a piece-wise function with two power-law segments jointed by a sudden drop. The other consists of a main truncated power-law component, a Gaussian component, and a third very massive component. Both models can reasonably fit the data and a sharp drop of the mass distribution is found at $\sim 50M_\odot$, suggesting that the majority of the observed black holes can be explained by the stellar evolution scenarios in which the pulsational pair-instability process takes place. On the other hand, the very massive sub-population, which accounts for at most several percents of the total, may be formed through hierarchical mergers or other processes.

astro-ph.HE↗

arXiv2021

Communication-Efficient Local Decentralized SGD Methods

Recently, the technique of local updates is a powerful tool in centralized settings to improve communication efficiency via periodical communication. For decentralized settings, it is still unclear how to efficiently combine local updates and decentralized communication. In this work, we propose an algorithm named as LD-SGD, which incorporates arbitrary update schemes that alternate between multiple Local updates and multiple Decentralized SGDs, and provide an analytical framework for LD-SGD. Under the framework, we present a sufficient condition to guarantee the convergence. We show that LD-SGD converges to a critical point for a wide range of update schemes when the objective is non-convex and the training data are non-identically independent distributed. Moreover, our framework brings many insights into the design of update schemes for decentralized optimization. As examples, we specify two update schemes and show how they help improve communication efficiency. Specifically, the first scheme alternates the number of local and global update steps. From our analysis, the ratio of the number of local updates to that of decentralized SGD trades off communication and computation. The second scheme is to periodically shrink the length of local updates. We show that the decaying strategy helps improve communication efficiency both theoretically and empirically.

stat.ML↗

arXiv2021

Development and Validation of a Deep Learning Model for Prediction of Severe Outcomes in Suspected COVID-19 Infection

COVID-19 patient triaging with predictive outcome of the patients upon first present to emergency department (ED) is crucial for improving patient prognosis, as well as better hospital resources management and cross-infection control. We trained a deep feature fusion model to predict patient outcomes, where the model inputs were EHR data including demographic information, co-morbidities, vital signs and laboratory measurements, plus patient's CXR images. The model output was patient outcomes defined as the most insensitive oxygen therapy required. For patients without CXR images, we employed Random Forest method for the prediction. Predictive risk scores for COVID-19 severe outcomes ("CO-RISK" score) were derived from model output and evaluated on the testing dataset, as well as compared to human performance. The study's dataset (the "MGB COVID Cohort") was constructed from all patients presenting to the Mass General Brigham (MGB) healthcare system from March 1st to June 1st, 2020. ED visits with incomplete or erroneous data were excluded. Patients with no test order for COVID or confirmed negative test results were excluded. Patients under the age of 15 were also excluded. Finally, electronic health record (EHR) data from a total of 11060 COVID-19 confirmed or suspected patients were used in this study. Chest X-ray (CXR) images were also collected from each patient if available. Results show that CO-RISK score achieved area under the Curve (AUC) of predicting MV/death (i.e. severe outcomes) in 24 hours of 0.95, and 0.92 in 72 hours on the testing dataset. The model shows superior performance to the commonly used risk scores in ED (CURB-65 and MEWS). Comparing with physician's decisions, CO-RISK score has demonstrated superior performance to human in making ICU/floor decisions.

cs.LG↗

arXiv2021

Bridging Unpaired Facial Photos And Sketches By Line-drawings

In this paper, we propose a novel method to learn face sketch synthesis models by using unpaired data. Our main idea is bridging the photo domain $\mathcal{X}$ and the sketch domain $Y$ by using the line-drawing domain $\mathcal{Z}$. Specially, we map both photos and sketches to line-drawings by using a neural style transfer method, i.e. $F: \mathcal{X}/\mathcal{Y} \mapsto \mathcal{Z}$. Consequently, we obtain \textit{pseudo paired data} $(\mathcal{Z}, \mathcal{Y})$, and can learn the mapping $G:\mathcal{Z} \mapsto \mathcal{Y}$ in a supervised learning manner. In the inference stage, given a facial photo, we can first transfer it to a line-drawing and then to a sketch by $G \circ F$. Additionally, we propose a novel stroke loss for generating different types of strokes. Our method, termed sRender, accords well with human artists' rendering process. Experimental results demonstrate that sRender can generate multi-style sketches, and significantly outperforms existing unpaired image-to-image translation methods.

cs.CV↗

arXiv2021

vrCAPTCHA: Exploring CAPTCHA Designs in Virtual Reality

With the popularity of online access in virtual reality (VR) devices, it will become important to investigate exclusive and interactive CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) designs for VR devices. In this paper, we first present four traditional two-dimensional (2D) CAPTCHAs (i.e., text-based, image-rotated, image-puzzled, and image-selected CAPTCHAs) in VR. Then, based on the three-dimensional (3D) interaction characteristics of VR devices, we propose two vrCAPTCHA design prototypes (i.e., task-driven and bodily motion-based CAPTCHAs). We conducted a user study with six participants for exploring the feasibility of our two vrCAPTCHAs and traditional CAPTCHAs in VR. We believe that our two vrCAPTCHAs can be an inspiration for the further design of CAPTCHAs in VR.

cs.HC↗

arXiv2021

Leveraging Meta-path Contexts for Classification in Heterogeneous Information Networks

A heterogeneous information network (HIN) has as vertices objects of different types and as edges the relations between objects, which are also of various types. We study the problem of classifying objects in HINs. Most existing methods perform poorly when given scarce labeled objects as training sets, and methods that improve classification accuracy under such scenarios are often computationally expensive. To address these problems, we propose ConCH, a graph neural network model. ConCH formulates the classification problem as a multi-task learning problem that combines semi-supervised learning with self-supervised learning to learn from both labeled and unlabeled data. ConCH employs meta-paths, which are sequences of object types that capture semantic relationships between objects. ConCH co-derives object embeddings and context embeddings via graph convolution. It also uses the attention mechanism to fuse such embeddings. We conduct extensive experiments to evaluate the performance of ConCH against other 15 classification methods. Our results show that ConCH is an effective and efficient method for HIN classification.

cs.LG↗

arXiv2021

Disentangling User Interest and Conformity for Recommendation with Causal Embedding

Recommendation models are usually trained on observational interaction data. However, observational interaction data could result from users' conformity towards popular items, which entangles users' real interest. Existing methods tracks this problem as eliminating popularity bias, e.g., by re-weighting training samples or leveraging a small fraction of unbiased data. However, the variety of user conformity is ignored by these approaches, and different causes of an interaction are bundled together as unified representations, hence robustness and interpretability are not guaranteed when underlying causes are changing. In this paper, we present DICE, a general framework that learns representations where interest and conformity are structurally disentangled, and various backbone recommendation models could be smoothly integrated. We assign users and items with separate embeddings for interest and conformity, and make each embedding capture only one cause by training with cause-specific data which is obtained according to the colliding effect of causal inference. Our proposed methodology outperforms state-of-the-art baselines with remarkable improvements on two real-world datasets on top of various backbone models. We further demonstrate that the learned embeddings successfully capture the desired causes, and show that DICE guarantees the robustness and interpretability of recommendation.

cs.IR↗

arXiv2021

SceneRec: Scene-Based Graph Neural Networks for Recommender Systems

Collaborative filtering has been largely used to advance modern recommender systems to predict user preference. A key component in collaborative filtering is representation learning, which aims to project users and items into a low dimensional space to capture collaborative signals. However, the scene information, which has effectively guided many recommendation tasks, is rarely considered in existing collaborative filtering methods. To bridge this gap, we focus on scene-based collaborative recommendation and propose a novel representation model SceneRec. SceneRec formally defines a scene as a set of pre-defined item categories that occur simultaneously in real-life situations and creatively designs an item-category-scene hierarchical structure to build a scene-based graph. In the scene-based graph, we adopt graph neural networks to learn scene-specific representation on each item node, which is further aggregated with latent representation learned from collaborative interactions to make recommendations. We perform extensive experiments on real-world E-commerce datasets and the results demonstrate the effectiveness of the proposed method.

cs.IR↗

arXiv2021

Contact Map based Crystal Structure Prediction using Global Optimization

Crystal structure prediction is now playing an increasingly important role in discovery of new materials. Global optimization methods such as genetic algorithms (GA) and particle swarm optimization (PSO) have been combined with first principle free energy calculations to predict crystal structures given composition or only a chemical system. While these approaches can exploit certain crystal patterns such as symmetry and periodicity in their search process, they usually do not exploit the large amount of implicit rules and constraints of atom configurations embodied in the large number of known crystal structures. They currently can only handle crystal structure prediction of relatively small systems. Inspired by the knowledge-rich protein structure prediction approach, herein we explore whether known geometric constraints such as the atomic contact map of a target crystal material can help predict its structure given its space group information. We propose a global optimization based algorithm, CMCrystal, for crystal structure reconstruction based on atomic contact maps. Based on extensive experiments using six global optimization algorithms, we show that it is viable to reconstruct the crystal structure given the atomic contact map for some crystal materials but more constraints are needed for other target materials to achieve successful reconstruction. This implies that atomic interaction information learned from existing materials can be used to improve crystal structure prediction.

cond-mat.mtrl-sci↗

arXiv2021

Emotion controllable speech synthesis using emotion-unlabeled dataset with the assistance of cross-domain speech emotion recognition

Neural text-to-speech (TTS) approaches generally require a huge number of high quality speech data, which makes it difficult to obtain such a dataset with extra emotion labels. In this paper, we propose a novel approach for emotional TTS synthesis on a TTS dataset without emotion labels. Specifically, our proposed method consists of a cross-domain speech emotion recognition (SER) model and an emotional TTS model. Firstly, we train the cross-domain SER model on both SER and TTS datasets. Then, we use emotion labels on the TTS dataset predicted by the trained SER model to build an auxiliary SER task and jointly train it with the TTS model. Experimental results show that our proposed method can generate speech with the specified emotional expressiveness and nearly no hindering on the speech quality.

eess.AS↗

arXiv2021

Towards Cross-Modal Forgery Detection and Localization on Live Surveillance Videos

The cybersecurity breaches render surveillance systems vulnerable to video forgery attacks, under which authentic live video streams are tampered to conceal illegal human activities under surveillance cameras. Traditional video forensics approaches can detect and localize forgery traces in each video frame using computationally-expensive spatial-temporal analysis, while falling short in real-time verification of live video feeds. The recent work correlates time-series camera and wireless signals to recognize replayed surveillance videos using event-level timing information but it cannot realize fine-grained forgery detection and localization on each frame. To fill this gap, this paper proposes Secure-Pose, a novel cross-modal forgery detection and localization system for live surveillance videos using WiFi signals near the camera spot. We observe that coexisting camera and WiFi signals convey common human semantic information and the presence of forgery attacks on video frames will decouple such information correspondence. Secure-Pose extracts effective human pose features from synchronized multi-modal signals and detects and localizes forgery traces under both inter-frame and intra-frame attacks in each frame. We implement Secure-Pose using a commercial camera and two Intel 5300 NICs and evaluate it in real-world environments. Secure-Pose achieves a high detection accuracy of 95.1% and can effectively localize tampered objects under different forgery attacks.

cs.CR↗

arXiv2021

Delayed Projection Techniques for Linearly Constrained Problems: Convergence Rates, Acceleration, and Applications

In this work, we study a novel class of projection-based algorithms for linearly constrained problems (LCPs) which have a lot of applications in statistics, optimization, and machine learning. Conventional primal gradient-based methods for LCPs call a projection after each (stochastic) gradient descent, resulting in that the required number of projections equals that of gradient descents (or total iterations). Motivated by the recent progress in distributed optimization, we propose the delayed projection technique that calls a projection once for a while, lowering the projection frequency and improving the projection efficiency. Accordingly, we devise a series of stochastic methods for LCPs using the technique, including a variance reduced method and an accelerated one. We theoretically show that it is feasible to improve projection efficiency in both strongly convex and generally convex cases. Our analysis is simple and unified and can be easily extended to other methods using delayed projections. When applying our new algorithms to federated optimization, a newfangled and privacy-preserving subfield in distributed optimization, we obtain not only a variance reduced federated algorithm with convergence rates better than previous works, but also the first accelerated method able to handle data heterogeneity inherent in federated optimization.

math.OC↗

arXiv2020

Observation of anti-damping spin-orbit torques generated by in-plane and out-of-plane spin polarizations in MnPd3

High spin-orbit torques (SOTs) generated by topological materials and heavy metals interfaced with a ferromagnetic layer show promise for next generation magnetic memory and logic devices. SOTs generated from the in-plane spin polarization along y-axis originated by the spin Hall and Edelstein effects can switch magnetization collinear with the spin polarization in the absence of external magnetic fields. However, an external magnetic field is required to switch the magnetization along x and z-axes via SOT generated by y-spin polarization. Here, we present that the above limitation can be circumvented by unconventional SOT in magnetron-sputtered thin film MnPd3. In addition to the conventional in-plane anti-damping-like torque due to the y-spin polarization, out-of-plane and in-plane anti-damping-like torques originating from z-spin and x-spin polarizations, respectively have been observed at room temperature. The spin torque efficiency corresponding to the y-spin polarization from MnPd3 thin films grown on thermally oxidized silicon substrate and post annealed at 400 Deg C is 0.34 - 0.44. Remarkably, we have demonstrated complete external magnetic field-free switching of perpendicular Co layer via unconventional out-of-plane anti-damping-like torque from z-spin polarization. Based on the density functional theory calculations, we determine that the observed x- and z- spin polarizations with the in-plane charge current are due to the low symmetry of the (114) oriented MnPd3 thin films. Taken together, the new material reported here provides a path to realize a practical spin channel in ultrafast magnetic memory and logic devices.

cond-mat.mtrl-sci↗

arXiv2020

Cost-Effective Federated Learning Design

Federated learning (FL) is a distributed learning paradigm that enables a large number of devices to collaboratively learn a model without sharing their raw data. Despite its practical efficiency and effectiveness, the iterative on-device learning process incurs a considerable cost in terms of learning time and energy consumption, which depends crucially on the number of selected clients and the number of local iterations in each training round. In this paper, we analyze how to design adaptive FL that optimally chooses these essential control variables to minimize the total cost while ensuring convergence. Theoretically, we analytically establish the relationship between the total cost and the control variables with the convergence upper bound. To efficiently solve the cost minimization problem, we develop a low-cost sampling-based algorithm to learn the convergence related unknown parameters. We derive important solution properties that effectively identify the design principles for different metric preferences. Practically, we evaluate our theoretical results both in a simulated environment and on a hardware prototype. Experimental evidence verifies our derived properties and demonstrates that our proposed solution achieves near-optimal performance for various datasets, different machine learning models, and heterogeneous system settings.

cs.LG↗

arXiv2020

Modeling Homophone Noise for Robust Neural Machine Translation

In this paper, we propose a robust neural machine translation (NMT) framework. The framework consists of a homophone noise detector and a syllable-aware NMT model to homophone errors. The detector identifies potential homophone errors in a textual sentence and converts them into syllables to form a mixed sequence that is then fed into the syllable-aware NMT. Extensive experiments on Chinese->English translation demonstrate that our proposed method not only significantly outperforms baselines on noisy test sets with homophone noise, but also achieves a substantial improvement on clean text.

cs.CL↗

arXiv2020

Simple model for sequential multiphoton ionization by ultra-intense x-rays

A simple model for sequential multiphoton ionization by ultra-intense x-rays is presented. The derived scaling of the ion yield with pulse energy quantitatively reproduces the experimental data, which shows that the ion yield increases according to the "power law" behavior typical of multiphoton ionization, followed by saturation at high pulse energies. The calculated average time interval between ionizations for producing ions at a certain charge state is found to be proportional to the pulse duration and independent of all other x-ray pulse parameters. This agrees with previous studies where the kinetic energy of fragment ions with a given charge state produced by intense x-ray ionization of molecules was found to be independent of the pulse energy, but to increase with smaller pulse duration due to the smaller time interval between ionizations.

physics.atom-ph↗