arXiv subjects

Xiang Li

Publications and source records attributed to Xiang Li.

At least 1,225 records · Page 68Linked to original sources

arXiv2020

Neural Image Compression and Explanation

Explaining the prediction of deep neural networks (DNNs) and semantic image compression are two active research areas of deep learning with a numerous of applications in decision-critical systems, such as surveillance cameras, drones and self-driving cars, where interpretable decision is critical and storage/network bandwidth is limited. In this paper, we propose a novel end-to-end Neural Image Compression and Explanation (NICE) framework that learns to (1) explain the predictions of convolutional neural networks (CNNs), and (2) subsequently compress the input images for efficient storage or transmission. Specifically, NICE generates a sparse mask over an input image by attaching a stochastic binary gate to each pixel of the image, whose parameters are learned through the interaction with the CNN classifier to be explained. The generated mask is able to capture the saliency of each pixel measured by its influence to the final prediction of CNN; it can also be used to produce a mixed-resolution image, where important pixels maintain their original high resolution and insignificant background pixels are subsampled to a low resolution. The produced images achieve a high compression rate (e.g., about 0.6x of original image file size), while retaining a similar classification accuracy. Extensive experiments across multiple image classification benchmarks demonstrate the superior performance of NICE compared to the state-of-the-art methods in terms of explanation quality and semantic image compression rate. Our code is available at: https://github.com/lxuniverse/NICE.

cs.CV↗

arXiv2020

V2I-Based Platooning Design with Delay Awareness

This paper studies the vehicle platooning system based on vehicle-to-infrastructure (V2I) communication, where all the vehicles in the platoon upload their driving state information to the roadside unit (RSU), and RSU makes the platoon control decisions with the assistance of edge computing. By addressing the delay concern, a platoon control approach is proposed to achieve plant stability and string stability. The effects of the time headway, communication and edge computing delays on the stability are quantified. The velocity and size of the stable platoon are calculated, which show the impacts of the radio parameters such as massive MIMO antennas and frequency band on the platoon configuration. The handover performance between RSUs in the V2I-based platooning system is quantified by considering the effects of the RSU's coverage and platoon size, which demonstrates that the velocity of a stable platoon should be appropriately chosen, in order to meet the V2I's Quality-of-Service and handover constraints.

cs.MA↗

arXiv2020

Broadband sensitivity improvement via coherent quantum feedback with PT symmetry

A conventional resonant detector is often subject to a trade-off between bandwidth and peak sensitivity that can be traced back to quantum Cramer-Rao Bound. Anomalous dispersion has been shown to improve it by signal amplification and is thus more robust against decoherence, while it leads to instabilities. We propose a stable quantum amplifier applicable to linear systems operating at the fundamental detection limits, enabled by two-mode non-degenerate parametric amplification. At threshold, one mode of the amplifier forms a PT-symmetric system of original detector mode. Sensitivity improvements are shown for laser-interferometric gravitational-wave detectors and microwave cavity axion detectors.

quant-ph↗

arXiv2020

Deep Metric Learning-based Image Retrieval System for Chest Radiograph and its Clinical Applications in COVID-19

In recent years, deep learning-based image analysis methods have been widely applied in computer-aided detection, diagnosis and prognosis, and has shown its value during the public health crisis of the novel coronavirus disease 2019 (COVID-19) pandemic. Chest radiograph (CXR) has been playing a crucial role in COVID-19 patient triaging, diagnosing and monitoring, particularly in the United States. Considering the mixed and unspecific signals in CXR, an image retrieval model of CXR that provides both similar images and associated clinical information can be more clinically meaningful than a direct image diagnostic model. In this work we develop a novel CXR image retrieval model based on deep metric learning. Unlike traditional diagnostic models which aims at learning the direct mapping from images to labels, the proposed model aims at learning the optimized embedding space of images, where images with the same labels and similar contents are pulled together. It utilizes multi-similarity loss with hard-mining sampling strategy and attention mechanism to learn the optimized embedding space, and provides similar images to the query image. The model is trained and validated on an international multi-site COVID-19 dataset collected from 3 different sources. Experimental results of COVID-19 image retrieval and diagnosis tasks show that the proposed model can serve as a robust solution for CXR analysis and patient management for COVID-19. The model is also tested on its transferability on a different clinical decision support task, where the pre-trained model is applied to extract image features from a new dataset without any further training. These results demonstrate our deep metric learning based image retrieval model is highly efficient in the CXR retrieval, diagnosis and prognosis, and thus has great clinical value for the treatment and management of COVID-19 patients.

eess.IV↗

arXiv2020

Energy-efficient Resource Allocation for Mobile Edge Computing Aided by Multiple Relays

In this paper, we study a mobile edge computing (MEC) system with the mobile device aided by multiple relay nodes for offloading data to an edge server. Specifically, the modes of decode-and-forward (DF) with time-division-multiple-access (TDMA) and frequency-division-multiple-access (FDMA), and the mode of amplify-and-forward (AF) are investigated, which are denoted as DF-TDMA, DF-FDMA, and AF, respectively. Our target is to minimize the total energy consumption of the mobile device and multiple relay nodes through optimizing the allocation of computation and communication resources. Optimization problems under the three considered modes are formulated and shown to be non-convex. For DF-TDMA mode, we transform the original non-convex problem to be a convex one and further develop a low computation complexity yet optimal solution. In DF-FDMA mode, with some transformation on the original problem, we prove the mathematical equivalence between the transformed problem in DF-FDMA mode and the problem under DF-TDMA mode. In AF mode, the associated optimization problem is decomposed into two levels, in which monotonic optimization is utilized in upper level and successive convex approximation (SCA) is adopted to find the convergent solution in the lower level. Numerical results prove the effectiveness of our proposed methods under various working modes.

cs.IT↗

arXiv2020

Generalized Focal Loss V2: Learning Reliable Localization Quality Estimation for Dense Object Detection

Localization Quality Estimation (LQE) is crucial and popular in the recent advancement of dense object detectors since it can provide accurate ranking scores that benefit the Non-Maximum Suppression processing and improve detection performance. As a common practice, most existing methods predict LQE scores through vanilla convolutional features shared with object classification or bounding box regression. In this paper, we explore a completely novel and different perspective to perform LQE -- based on the learned distributions of the four parameters of the bounding box. The bounding box distributions are inspired and introduced as "General Distribution" in GFLV1, which describes the uncertainty of the predicted bounding boxes well. Such a property makes the distribution statistics of a bounding box highly correlated to its real localization quality. Specifically, a bounding box distribution with a sharp peak usually corresponds to high localization quality, and vice versa. By leveraging the close correlation between distribution statistics and the real localization quality, we develop a considerably lightweight Distribution-Guided Quality Predictor (DGQP) for reliable LQE based on GFLV1, thus producing GFLV2. To our best knowledge, it is the first attempt in object detection to use a highly relevant, statistical representation to facilitate LQE. Extensive experiments demonstrate the effectiveness of our method. Notably, GFLV2 (ResNet-101) achieves 46.2 AP at 14.6 FPS, surpassing the previous state-of-the-art ATSS baseline (43.6 AP at 14.6 FPS) by absolute 2.6 AP on COCO {\tt test-dev}, without sacrificing the efficiency both in training and inference. Code will be available at https://github.com/implus/GFocalV2.

cs.CV↗

arXiv2020

Retrieving and ranking short medical questions with two stages neural matching model

Internet hospital is a rising business thanks to recent advances in mobile web technology and high demand of health care services. Online medical services become increasingly popular and active. According to US data in 2018, 80 percent of internet users have asked health-related questions online. Numerous data is generated in unprecedented speed and scale. Those representative questions and answers in medical fields are valuable raw data sources for medical data mining. Automated machine interpretation on those sheer amount of data gives an opportunity to assist doctors to answer frequently asked medical-related questions from the perspective of information retrieval and machine learning approaches. In this work, we propose a novel two-stage framework for the semantic matching of query-level medical questions.

cs.CL↗

arXiv2020

Finding the Near Optimal Policy via Adaptive Reduced Regularization in MDPs

Regularized MDPs serve as a smooth version of original MDPs. However, biased optimal policy always exists for regularized MDPs. Instead of making the coefficientλof regularized term sufficiently small, we propose an adaptive reduction scheme for λ to approximate optimal policy of the original MDP. It is shown that the iteration complexity for obtaining anε-optimal policy could be reduced in comparison with setting sufficiently smallλ. In addition, there exists strong duality connection between the reduction method and solving the original MDP directly, from which we can derive more adaptive reduction method for certain algorithms.

cs.LG↗

arXiv2020

3D Meta-Registration: Learning to Learn Registration of 3D Point Clouds

Deep learning-based point cloud registration models are often generalized from extensive training over a large volume of data to learn the ability to predict the desired geometric transformation to register 3D point clouds. In this paper, we propose a meta-learning based 3D registration model, named 3D Meta-Registration, that is capable of rapidly adapting and well generalizing to new 3D registration tasks for unseen 3D point clouds. Our 3D Meta-Registration gains a competitive advantage by training over a variety of 3D registration tasks, which leads to an optimized model for the best performance on the distribution of registration tasks including potentially unseen tasks. Specifically, the proposed 3D Meta-Registration model consists of two modules: 3D registration learner and 3D registration meta-learner. During the training, the 3D registration learner is trained to complete a specific registration task aiming to determine the desired geometric transformation that aligns the source point cloud with the target one. In the meantime, the 3D registration meta-learner is trained to provide the optimal parameters to update the 3D registration learner based on the learned task distribution. After training, the 3D registration meta-learner, which is learned with the optimized coverage of distribution of 3D registration tasks, is able to dynamically update 3D registration learners with desired parameters to rapidly adapt to new registration tasks. We tested our model on synthesized dataset ModelNet and FlyingThings3D, as well as real-world dataset KITTI. Experimental results demonstrate that 3D Meta-Registration achieves superior performance over other previous techniques (e.g. FlowNet3D).

cs.CV↗

arXiv2020

3D Meta Point Signature: Learning to Learn 3D Point Signature for 3D Dense Shape Correspondence

Point signature, a representation describing the structural neighborhood of a point in 3D shapes, can be applied to establish correspondences between points in 3D shapes. Conventional methods apply a weight-sharing network, e.g., any kind of graph neural networks, across all neighborhoods to directly generate point signatures and gain the generalization ability by extensive training over a large amount of training samples from scratch. However, these methods lack the flexibility in rapidly adapting to unseen neighborhood structures and thus generalizes poorly on new point sets. In this paper, we propose a novel meta-learning based 3D point signature model, named 3Dmetapointsignature (MEPS) network, that is capable of learning robust point signatures in 3D shapes. By regarding each point signature learning process as a task, our method obtains an optimized model over the best performance on the distribution of all tasks, generating reliable signatures for new tasks, i.e., signatures of unseen point neighborhoods. Specifically, the MEPS consists of two modules: a base signature learner and a meta signature learner. During training, the base-learner is trained to perform specific signature learning tasks. In the meantime, the meta-learner is trained to update the base-learner with optimal parameters. During testing, the meta-learner that is learned with the distribution of all tasks can adaptively change parameters of the base-learner, accommodating to unseen local neighborhoods. We evaluate the MEPS model on two datasets, e.g., FAUST and TOSCA, for dense 3Dshape correspondence. Experimental results demonstrate that our method not only gains significant improvements over the baseline model and achieves state-of-the-art results, but also is capable of handling unseen 3D shapes.

cs.CV↗

arXiv2020

A generalized Catoni's ${\rm M}$-estimator under finite {$α$-th moment assumption} with $α\in (1,2)$

We generalize the { ${\rm M}$-estimator} put forward by Catoni in his seminal paper [C12] to the case in which samples can have finite $α$-th moment with $α\in (1,2)$ rather than finite variance, our approach is by slightly modifying the influence function $φ$ therein. The choice of the new influence function is inspired by the Taylor-like expansion developed in [C-N-X]. We obtain a deviation bound of the estimator, as $α\rightarrow 2$, this bound is the same as that in [C12]. Experiment shows that our generalized ${\rm M}$-estimator performs better than the empirical mean estimator, the smaller the $α$ is, the better the performance will be. As an application, we study an $\ell_{1}$ regression considered by Zhang et al. [Z-Z] who assumed that samples have finite variance, and relax their assumption to be finite {$α$-th} moment with $α\in (1,2)$.

math.ST↗

arXiv2020

VirusBoxing: A HIIT-based VR boxing game

Physical activity or exercise can improve people's health and reduce their risk of developing several diseases; most importantly, regular activity can improve the quality of life. However, lack of time is one of the major barriers for people doing exercise. High-intensity interval training (HIIT) can reduce the time required for a healthy exercise regime but also bring similar benefits of regular exercise. We present a boxing-based VR exergame called VirusBoxing to promote physical activity for players. VirusBoxing provides players with a platform for HIIT and empowers them with additional abilities to jab a distant object without the need to aim at it precisely. In this paper, we discuss how we adapted the HIIT protocol and gameplay features to empower players in a VR exergame to give players an efficient, effective, and enjoyable exercise experience.

cs.HC↗

arXiv2020

Exploration of Hands-free Text Entry Techniques For Virtual Reality

Text entry is a common activity in virtual reality (VR) systems. There is a limited number of available hands-free techniques, which allow users to carry out text entry when users' hands are busy such as holding items or hand-based devices are not available. The most used hands-free text entry technique is DwellType, where a user selects a letter by dwelling over it for a specific period. However, its performance is limited due to the fixed dwell time for each character selection. In this paper, we explore two other hands-free text entry mechanisms in VR: BlinkType and NeckType, which leverage users' eye blinks and neck's forward and backward movements to select letters. With a user study, we compare the performance of the two techniques with DwellType. Results show that users can achieve an average text entry rate of 13.47, 11.18 and 11.65 words per minute with BlinkType, NeckType, and DwellType, respectively. Users' subjective feedback shows BlinkType as the preferred technique for text entry in VR.

cs.HC↗

arXiv2020

Reading Comprehension as Natural Language Inference: A Semantic Analysis

In the recent past, Natural language Inference (NLI) has gained significant attention, particularly given its promise for downstream NLP tasks. However, its true impact is limited and has not been well studied. Therefore, in this paper, we explore the utility of NLI for one of the most prominent downstream tasks, viz. Question Answering (QA). We transform the one of the largest available MRC dataset (RACE) to an NLI form, and compare the performances of a state-of-the-art model (RoBERTa) on both these forms. We propose new characterizations of questions, and evaluate the performance of QA and NLI models on these categories. We highlight clear categories for which the model is able to perform better when the data is presented in a coherent entailment form, and a structured question-answer concatenation form, respectively.

cs.CL↗

arXiv2020

Deep-3DAligner: Unsupervised 3D Point Set Registration Network With Optimizable Latent Vector

Point cloud registration is the process of aligning a pair of point sets via searching for a geometric transformation. Unlike classical optimization-based methods, recent learning-based methods leverage the power of deep learning for registering a pair of point sets. In this paper, we propose to develop a novel model that organically integrates the optimization to learning, aiming to address the technical challenges in 3D registration. More specifically, in addition to the deep transformation decoding network, our framework introduce an optimizable deep \underline{S}patial \underline{C}orrelation \underline{R}epresentation (SCR) feature. The SCR feature and weights of the transformation decoder network are jointly updated towards the minimization of an unsupervised alignment loss. We further propose an adaptive Chamfer loss for aligning partial shapes. To verify the performance of our proposed method, we conducted extensive experiments on the ModelNet40 dataset. The results demonstrate that our method achieves significantly better performance than the previous state-of-the-art approaches in the full/partial point set registration task.

cs.CV↗

arXiv2020

Comparison of proton shower developments in the BGO calorimeter of the Dark Matter Particle Explorer between GEANT4 and FLUKA simulations

The DArk Matter Particle Explorer (DAMPE) is a satellite-borne detector for high-energy cosmic rays and $γ$-rays. To fully understand the detector performance and obtain reliable physical results, extensive simulations of the detector are necessary. The simulations are particularly important for the data analysis of cosmic ray nuclei, which relies closely on the hadronic and nuclear interactions of particles in the detector material. Widely adopted simulation softwares include the GEANT4 and FLUKA, both of which have been implemented for the DAMPE simulation tool. Here we describe the simulation tool of DAMPE and compare the results of proton shower properties in the calorimeter from the two simulation softwares. Such a comparison gives an estimate of the most significant uncertainties of our proton spectral analysis.

physics.ins-det↗

arXiv2020

Geometry-Aware Segmentation of Remote Sensing Images via Implicit Height Estimation

Recent studies have shown the benefits of using additional elevation data (e.g., DSM) for enhancing the performance of the semantic segmentation of aerial images. However, previous methods mostly adopt 3D elevation information as additional inputs. While in many real-world applications, one does not have the corresponding DSM information at hand and the spatial resolution of acquired DSM images usually do not match the aerial images. To alleviate this data constraint and also take advantage of 3D elevation information, in this paper, we introduce a geometry-aware segmentation model that achieves accurate semantic labeling of aerial images via joint height estimation. Instead of using a single-stream encoder-decoder network for semantic labeling, we design a separate decoder branch to predict the height map and use the DSM images as side supervision to train this newly designed decoder branch. In this way, our model does not require DSM as model input and still benefits from the helpful 3D geometric information during training. Moreover, we develop a new geometry-aware convolution module that fuses the 3D geometric features from the height decoder branch and the 2D contextual features from the semantic segmentation branch. The fused feature embeddings can produce geometry-aware segmentation maps with enhanced performance. Our model is trained with DSM images as side supervision, while in the inference stage, it does not require DSM data and directly predicts the semantic labels in an end-to-end fashion. Experiments on ISPRS Vaihingen and Potsdam datasets demonstrate the effectiveness of the proposed method for the semantic segmentation of aerial images. The proposed model achieves remarkable performance on both datasets without using any hand-crafted features or post-processing.

cs.CV↗

arXiv2020

Online Speaker Diarization with Relation Network

In this paper, we propose an online speaker diarization system based on Relation Network, named RenoSD. Unlike conventional diariztion systems which consist of several independently-optimized modules, RenoSD implements voice-activity-detection (VAD), embedding extraction, and speaker identity association using a single deep neural network. The most striking feature of RenoSD is that it adopts a meta-learning strategy for speaker identity association. In particular, the relation network learns to learn a deep distance metric in a data-driven way and it can determine through a simple forward pass whether two given segments belong to the same speaker. As such, RenoSD can be performed in an online manner with low latency. Experimental results on AMI and CALLHOME datasets show that the proposed RenoSD system achieves consistent improvements over the state-of-the-art x-vector baseline. Compared with an existing online diarization system named UIS-RNN, RenoSD achieves a better performance using much fewer training data and at a lower time complexity.

eess.AS↗