arXiv subjects

Ye Yuan

Publications and source records attributed to Ye Yuan.

At least 289 records · Page 16Linked to original sources

arXiv2020

UG$^{2+}$ Track 2: A Collective Benchmark Effort for Evaluating and Advancing Image Understanding in Poor Visibility Environments

The UG$^{2+}$ challenge in IEEE CVPR 2019 aims to evoke a comprehensive discussion and exploration about how low-level vision techniques can benefit the high-level automatic visual recognition in various scenarios. In its second track, we focus on object or face detection in poor visibility enhancements caused by bad weathers (haze, rain) and low light conditions. While existing enhancement methods are empirically expected to help the high-level end task, that is observed to not always be the case in practice. To provide a more thorough examination and fair comparison, we introduce three benchmark sets collected in real-world hazy, rainy, and low-light conditions, respectively, with annotate objects/faces annotated. To our best knowledge, this is the first and currently largest effort of its kind. Baseline results by cascading existing enhancement and detection models are reported, indicating the highly challenging nature of our new data as well as the large room for further technical innovations. We expect a large participation from the broad research community to address these challenges together.

cs.CV↗

arXiv2020

Optical Non-Line-of-Sight Physics-based 3D Human Pose Estimation

We describe a method for 3D human pose estimation from transient images (i.e., a 3D spatio-temporal histogram of photons) acquired by an optical non-line-of-sight (NLOS) imaging system. Our method can perceive 3D human pose by `looking around corners' through the use of light indirectly reflected by the environment. We bring together a diverse set of technologies from NLOS imaging, human pose estimation and deep reinforcement learning to construct an end-to-end data processing pipeline that converts a raw stream of photon measurements into a full 3D human pose sequence estimate. Our contributions are the design of data representation process which includes (1) a learnable inverse point spread function (PSF) to convert raw transient images into a deep feature vector; (2) a neural humanoid control policy conditioned on the transient image feature and learned from interactions with a physics simulator; and (3) a data synthesis and augmentation strategy based on depth data that can be transferred to a real-world NLOS imaging system. Our preliminary experiments suggest that our method is able to generalize to real-world NLOS measurement to estimate physically-valid 3D human poses.

cs.CV↗

arXiv2020

Uncertainty Quantification for Deep Context-Aware Mobile Activity Recognition and Unknown Context Discovery

Activity recognition in wearable computing faces two key challenges: i) activity characteristics may be context-dependent and change under different contexts or situations; ii) unknown contexts and activities may occur from time to time, requiring flexibility and adaptability of the algorithm. We develop a context-aware mixture of deep models termed the α-\b{eta} network coupled with uncertainty quantification (UQ) based upon maximum entropy to enhance human activity recognition performance. We improve accuracy and F score by 10% by identifying high-level contexts in a data-driven way to guide model development. In order to ensure training stability, we have used a clustering-based pre-training in both public and in-house datasets, demonstrating improved accuracy through unknown context discovery.

cs.LG↗

arXiv2020

State-Aware Tracker for Real-Time Video Object Segmentation

In this work, we address the task of semi-supervised video object segmentation(VOS) and explore how to make efficient use of video property to tackle the challenge of semi-supervision. We propose a novel pipeline called State-Aware Tracker(SAT), which can produce accurate segmentation results with real-time speed. For higher efficiency, SAT takes advantage of the inter-frame consistency and deals with each target object as a tracklet. For more stable and robust performance over video sequences, SAT gets awareness for each state and makes self-adaptation via two feedback loops. One loop assists SAT in generating more stable tracklets. The other loop helps to construct a more robust and holistic target representation. SAT achieves a promising result of 72.3% J&F mean with 39 FPS on DAVIS2017-Val dataset, which shows a decent trade-off between efficiency and accuracy. Code will be released at github.com/MegviiDetection/video_analyst.

cs.CV↗

arXiv2020

Bridging the Gap Between Computational Photography and Visual Recognition

What is the current state-of-the-art for image restoration and enhancement applied to degraded images acquired under less than ideal circumstances? Can the application of such algorithms as a pre-processing step to improve image interpretability for manual analysis or automatic visual recognition to classify scene content? While there have been important advances in the area of computational photography to restore or enhance the visual quality of an image, the capabilities of such techniques have not always translated in a useful way to visual recognition tasks. Consequently, there is a pressing need for the development of algorithms that are designed for the joint problem of improving visual appearance and recognition, which will be an enabling factor for the deployment of visual recognition tools in many real-world scenarios. To address this, we introduce the UG^2 dataset as a large-scale benchmark composed of video imagery captured under challenging conditions, and two enhancement tasks designed to test algorithmic impact on visual quality and automatic object recognition. Furthermore, we propose a set of metrics to evaluate the joint improvement of such tasks as well as individual algorithmic advances, including a novel psychophysics-based evaluation regime for human assessment and a realistic set of quantitative measures for object recognition performance. We introduce six new algorithms for image restoration or enhancement, which were created as part of the IARPA sponsored UG^2 Challenge workshop held at CVPR 2018. Under the proposed evaluation regime, we present an in-depth analysis of these algorithms and a host of deep learning-based and classic baseline approaches. From the observed results, it is evident that we are in the early days of building a bridge between computational photography and visual recognition, leaving many opportunities for innovation in this area.

cs.CV↗

arXiv2020

Colloidal interactions and unusual crystallization versus de-mixing of elastic multipoles formed by gold mesoflowers

Colloidal interactions in nematic liquid crystals can be described as interactions between elastic multipoles that depend on particle shape, topology, chirality, boundary conditions and induced topological defects. Here, we describe a nematic colloidal system consisting of mesostructures of gold capable of inducing elastic multipoles of different order. Elastic monopoles are formed by relatively large asymmetric mesoflower particles, for which gravity and elastic torque balancing yields monopole-type interactions. High-order multipoles are instead formed by smaller mesoflowers with a myriad of shapes corresponding to multipoles of different orders, consistent with our computer simulations based on free energy minimization. We reveal unexpected many-body interactions in this colloidal system, ranging from de-mixing of elastic monopoles to a zoo of unusual colloidal crystals formed by high-order multipoles like hexadecapoles. Our findings show that gold mesoflowers may serve as a designer toolkit for engineering colloidal interaction and self-assembly, potentially exceeding that in atomic and molecular systems.

cond-mat.soft↗

arXiv2020

Elastic colloidal monopoles and reconfigurable self-assembly in liquid crystals

Monopole-like electrostatic interactions are ubiquitous in biology and condensed matter, but they are often screened by counter-ions and cannot be switched from attractive to repulsive. In colloidal science, where the prime goal is to develop colloidal particles that mimic and exceed the diversity and length-scales of atomic and molecular assembly, electrostatically charged particles cannot change the sign of their surface charge or transform from monopoles to higher-order multipoles. In liquid-crystal colloids, elastic interactions between particles arise to minimize the free energy associated with elastic distortions in the long-range alignment of rod-like molecules around the particles. In dipolar, quadrupolar and hexadecapolar nematic colloids, the symmetries of such elastic distortions mimic both electrostatic multipoles and the outmost occupied electron shells of atoms. Electric and magnetic switching, spontaneous transformations and optical control of elastic multipoles, as well as their interactions with topological defects and surface boundary conditions, have been demonstrated in such colloids. However, it has long been understood that elastic monopoles should relax to uniform or higher-order multipole states because of the elastic torques that they induce. Here we develop nematic colloids with strong elastic monopole moments and with elastic torques balanced by optical torques exerted by ambient light. We demonstrate the monopole-to-quadrupole reconfiguration of these colloidal particles by unstructured light, which resembles the driving of atoms between the ground state and various excited states. We show that the sign of the elastic monopoles can be switched, and that like-charged monopoles attract whereas oppositely charged ones repel, unlike in electrostatics. We also demonstrate the out-of-equilibrium dynamic assembly of these colloidal particles.

cond-mat.soft↗

arXiv2019

High Precision Variational Bayesian Inference of Sparse Linear Networks

Sparse networks can be found in a wide range of applications, such as biological and communication networks. Inference of such networks from data has been receiving considerable attention lately, mainly driven by the need to understand and control internal working mechanisms. However, while most available methods have been successful at predicting many correct links, they also tend to infer many incorrect links. Precision is the ratio between the number of correctly inferred links and all inferred links, and should ideally be close to 100%. For example, 50% precision means that half of inferred links are incorrect, and there is only a 50% chance of picking a correct one. In contrast, this paper develops a method, based on variational Bayesian inference and Gaussian processes, that focuses on inferring links with very high precision. In addition, our method does not require full-state measurements and effectively promotes both system stability and network sparsity. Monte Carlo simulations illustrate that our method has 100% or nearly 100% precision, even in the presence of noise. The method should be applicable to a wide range of network inference contexts, including biological networks and power systems.

eess.SY↗

arXiv2019

Diverse Trajectory Forecasting with Determinantal Point Processes

The ability to forecast a set of likely yet diverse possible future behaviors of an agent (e.g., future trajectories of a pedestrian) is essential for safety-critical perception systems (e.g., autonomous vehicles). In particular, a set of possible future behaviors generated by the system must be diverse to account for all possible outcomes in order to take necessary safety precautions. It is not sufficient to maintain a set of the most likely future outcomes because the set may only contain perturbations of a single outcome. While generative models such as variational autoencoders (VAEs) have been shown to be a powerful tool for learning a distribution over future trajectories, randomly drawn samples from the learned implicit likelihood model may not be diverse -- the likelihood model is derived from the training data distribution and the samples will concentrate around the major mode that has most data. In this work, we propose to learn a diversity sampling function (DSF) that generates a diverse and likely set of future trajectories. The DSF maps forecasting context features to a set of latent codes which can be decoded by a generative model (e.g., VAE) into a set of diverse trajectory samples. Concretely, the process of identifying the diverse set of samples is posed as a parameter estimation of the DSF. To learn the parameters of the DSF, the diversity of the trajectory samples is evaluated by a diversity loss based on a determinantal point process (DPP). Gradient descent is performed over the DSF parameters, which in turn move the latent codes of the sample set to find an optimal diverse and likely set of trajectories. Our method is a novel application of DPPs to optimize a set of items (trajectories) in continuous space. We demonstrate the diversity of the trajectories produced by our approach on both low-dimensional 2D trajectory data and high-dimensional human motion data.

cs.CV↗

arXiv2019

In Defense of the Triplet Loss Again: Learning Robust Person Re-Identification with Fast Approximated Triplet Loss and Label Distillation

The comparative losses (typically, triplet loss) are appealing choices for learning person re-identification (ReID) features. However, the triplet loss is computationally much more expensive than the (practically more popular) classification loss, limiting their wider usage in massive datasets. Moreover, the abundance of label noise and outliers in ReID datasets may also put the margin-based loss in jeopardy. This work addresses the above two shortcomings of triplet loss, extending its effectiveness to large-scale ReID datasets with potentially noisy labels. We propose a fast-approximated triplet (FAT) loss, which provably converts the point-wise triplet loss into its upper bound form, consisting of a point-to-set loss term plus cluster compactness regularization. It preserves the effectiveness of triplet loss, while leading to linear complexity to the training set size. A label distillation strategy is further designed to learn refined soft-labels in place of the potentially noisy labels, from only an identified subset of confident examples, through teacher-student networks. We conduct extensive experiments on three most popular ReID benchmarks (Market-1501, DukeMTMC-reID, and MSMT17), and demonstrate that FAT loss with distilled labels lead to ReID features with remarkable accuracy, efficiency, robustness, and direct transferability to unseen datasets.

cs.CV↗

arXiv2019

Optimize TSK Fuzzy Systems for Regression Problems: Mini-Batch Gradient Descent with Regularization, DropRule and AdaBound (MBGD-RDA)

Takagi-Sugeno-Kang (TSK) fuzzy systems are very useful machine learning models for regression problems. However, to our knowledge, there has not existed an efficient and effective training algorithm that ensures their generalization performance, and also enables them to deal with big data. Inspired by the connections between TSK fuzzy systems and neural networks, we extend three powerful neural network optimization techniques, i.e., mini-batch gradient descent, regularization, and AdaBound, to TSK fuzzy systems, and also propose three novel techniques (DropRule, DropMF, and DropMembership) specifically for training TSK fuzzy systems. Our final algorithm, mini-batch gradient descent with regularization, DropRule and AdaBound (MBGD-RDA), can achieve fast convergence in training TSK fuzzy systems, and also superior generalization performance in testing. It can be used for training TSK fuzzy systems on datasets of any size; however, it is particularly useful for big datasets, on which currently no other efficient training algorithms exist.

cs.LG↗

arXiv2019

Calibrated Domain-Invariant Learning for Highly Generalizable Large Scale Re-Identification

Many real-world applications, such as city-scale traffic monitoring and control, requires large-scale re-identification. However, previous ReID methods often failed to address two limitations in existing ReID benchmarks, i.e., low spatiotemporal coverage and sample imbalance. Notwithstanding their demonstrated success in every single benchmark, they have difficulties in generalizing to unseen environments. As a result, these methods are less applicable in a large-scale setting due to poor generalization. In seek for a highly generalizable large-scale ReID method, we present an adversarial domain invariant feature learning framework (ADIN) that explicitly learns to separate identity-related features from challenging variations, where for the first time "free" annotations in ReID data such as video timestamp and camera index are utilized. Furthermore, we find that the imbalance of nuisance classes jeopardizes the adversarial training, and for mitigation we propose a calibrated adversarial loss that is attentive to nuisance distribution. Experiments on existing large-scale person vehicle ReID datasets demonstrate that ADIN learns more robust and generalizable representations, as evidenced by its outstanding direct transfer performance across datasets, which is a criterion that can better measure the generalizability of large-scale ReID methods/

cs.CV↗

arXiv2019

Cracking In-Memory Database Index A Case Study for Adaptive Radix Tree Index

Indexes provide a method to access data in databases quickly. It can improve the response speed of subsequent queries by building a complete index in advance. However, it also leads to a huge overhead of the continuous updating during creating the index. An in-memory database usually has a higher query processing performance than disk databases and is more suitable for real-time query processing. Therefore, there is an urgent need to reduce the index creation and update cost for in-memory databases. Database cracking technology is currently recognized as an effective method to reduce the index initialization time. However, conventional cracking algorithms are focused on simple column data structure rather than those complex index structure for in-memory databases. In order to show the feasibility of in-memory database index cracking and promote to future more extensive research, this paper conducted a case study on the Adaptive Radix Tree (ART), a popular tree index structure of in-memory databases. On the basis of carefully examining the ART index construction overhead, an algorithm using auxiliary data structures to crack the ART index is proposed.

cs.DB↗

arXiv2019

Mining Bursting Communities in Temporal Graphs

Temporal graphs are ubiquitous. Mining communities that are bursting in a period of time is essential to seek emergency events in temporal graphs. Unfortunately, most previous studies for community mining in temporal networks ignore the bursting patterns of communities. In this paper, we are the first to study a problem of seeking bursting communities in a temporal graph. We propose a novel model, called (l, δ)-maximal dense core, to represent a bursting community in a temporal graph. Specifically, an (l, δ)-maximal dense core is a temporal subgraph in which each node has average degree no less than δ in a time segment with length no less than l. To compute the (l, δ)-maximal dense core, we first develop a novel dynamic programming algorithm which can calculate the segment density efficiently. Then, we propose an improved algorithm with several novel pruning techniques to further improve the efficiency. In addition, we also develop an efficient algorithm to enumerate all (l, δ)-maximal dense cores that are not dominated by the others in terms of the parameters l and δ. The results of extensive experiments on 9 real-life datasets demonstrate the effectiveness, efficiency and scalability of our algorithms.

cs.SI↗

arXiv2019

Jammed packings of 3D superellipsoids with tunable packing fraction, contact number, and ordering

We carry out numerical studies of static packings of frictionless superellipsoidal particles in three spatial dimensions. We consider more than $200$ different particle shapes by varying the three shape parameters that define superellipsoids. We characterize the structural and mechanical properties of both disordered and ordered packings using two packing-generation protocols. We perform athermal quasi-static compression simulations starting from either random, dilute configurations (Protocol 1) or thermalized, dense configurations (protocol $2$), which allows us to tune the orientational order of the packings. In general, we find that the contact numbers at jamming onset for superellipsoid packings are hypostatic, with $z_J < z_{\rm iso}$, where $z_{\rm iso} = 2d_f$ and $d_f = 5$ or $6$ depending on whether the particles are axi-symmetric or not. Over the full range of orientational order, we find that the number of quartic modes of the dynamical matrix for the packings always matches the number of missing contacts relative to the isostatic value. This result suggests that there are no mechanically redundant contacts for ordered, yet hypostatic packings of superellipsoidal particles. Additionally, we find that the packing fraction at jamming onset for disordered packings of superellipsoidal particles can be collapsed using two particle shape parameters, e.g. the asphericity $\mathcal{A}$ and reduced aspect ratio $β$ of the particles.

cond-mat.soft↗

arXiv2019

Machine Discovery of Partial Differential Equations from Spatiotemporal Data

The study presents a general framework for discovering underlying Partial Differential Equations (PDEs) using measured spatiotemporal data. The method, called Sparse Spatiotemporal System Discovery ($\text{S}^3\text{d}$), decides which physical terms are necessary and which can be removed (because they are physically negligible in the sense that they do not affect the dynamics too much) from a pool of candidate functions. The method is built on the recent development of Sparse Bayesian Learning; which enforces the sparsity in the to-be-identified PDEs, and therefore can balance the model complexity and fitting error with theoretical guarantees. Without leveraging prior knowledge or assumptions in the discovery process, we use an automated approach to discover ten types of PDEs, including the famous Navier-Stokes and sine-Gordon equations, from simulation data alone. Moreover, we demonstrate our data-driven discovery process with the Complex Ginzburg-Landau Equation (CGLE) using data measured from a traveling-wave convection experiment. Our machine discovery approach presents solutions that has the potential to inspire, support and assist physicists for the establishment of physical laws from measured spatiotemporal data, especially in notorious fields that are often too complex to allow a straightforward establishment of physical law, such as biophysics, fluid dynamics, neuroscience or nonlinear optics.

stat.ML↗

arXiv2019

p-type codoping effect in (Ga,Mn)As: Mn lattice location versus magnetic properties

In the present work, we perform a systematic investigation on p-type codoping in (Ga,Mn)As. Through gradually increasing Zn doping concentration, the hole concentration increases, which should theoretically lead to an increase of the Curie temperature (TC) according to the p-d Zener model. Unexpectedly, although the film keeps its epitaxial structure, both TC and the magnetization decrease. The samples present a phase transition from ferromagnetism to paramagnetism upon increasing hole concentration. In the intermediate regime, we observe a signature of antiferromagnetism. By using channeling Rutherford backscattering spectrometry and particle-induced x-ray emission, the substitutional Mn atoms are observed to shift to interstitial sites, while more Zn atoms occupy Ga sites, which explains the observed behavior. This is also consistent with first-principles calculations, showing that the complex of substitutional Zn and interstitial Mn has the lowest formation energy.

cond-mat.mtrl-sci↗

arXiv2019

Blind SAR Image Despeckling Using Self-Supervised Dense Dilated Convolutional Neural Network

Despeckling is a key and indispensable step in SAR image preprocessing, existing deep learning-based methods achieve SAR despeckling by learning some mappings between speckled (different looks) and clean images. However, there exist no clean SAR image in the real world. To this end, in this paper, we propose a self-supervised dense dilated convolutional neural network (BDSS) for blind SAR image despeckling. Proposed BDSS can still learn to suppress speckle noise without clean ground truth by optimized for L2 loss. Besides, three enhanced dense blocks with dilated convolution are employed to improve network performance. The synthetic and real-data experiments demonstrate that proposed BDSS can achieve despeckling effectively while maintaining well features such as edges, point targets, and radiometric. At last, we demonstrate that our proposed BDSS can achieve blind despeckling excellently, i.e., do not need to care about the number of looks.

eess.IV↗