arXiv subjects

Zhe Chen

Publications and source records attributed to Zhe Chen.

At least 271 records · Page 15Linked to original sources

arXiv2021

Cosmology of the companion-axion model: dark matter, gravitational waves, and primordial black holes

The companion-axion model introduces a second QCD axion to rescue the Peccei-Quinn solution to the strong-CP problem from the effects of colored gravitational instantons. As in single-axion models, the two axions predicted by the companion-axion model are attractive candidates for dark matter. The model is defined by two free parameters, the scales of the two axions' symmetry breaking, so we can classify production scenarios in terms of the relative sizes of these two scales with respect to the scale of inflation. We study the cosmological production of companion-axion dark matter in order to predict windows of preferred axion masses, and calculate the relative abundances of the two particles. Additionally, we show that the presence of a second axion solves the cosmological domain wall problem automatically in the scenarios in which one or both of the axions are post-inflationary. We also suggest unique cosmological signatures of the companion-axion model, such as the production of a $\sim$10 nHz gravitational wave background, and $\sim 100\,M_\odot$ primordial black holes.

hep-ph↗

arXiv2021

Phenomenology of the companion-axion model: photon couplings

We study the phenomenology of the 'companion-axion model' consisting of two coupled QCD axions. The second axion is required to rescue the Peccei-Quinn solution to the strong-CP problem from the effects of colored gravitational instantons. We investigate here the combined phenomenology of axion-axion and axion-photon interactions, recasting present and future single-axion bounds onto the companion-axion parameter space. Most remarkably, we predict that future axion searches with haloscopes and helioscopes may well discover two QCD axions, perhaps even within the same experiment.

hep-ph↗

arXiv2021

Expression Snippet Transformer for Robust Video-based Facial Expression Recognition

The recent success of Transformer has provided a new direction to various visual understanding tasks, including video-based facial expression recognition (FER). By modeling visual relations effectively, Transformer has shown its power for describing complicated patterns. However, Transformer still performs unsatisfactorily to notice subtle facial expression movements, because the expression movements of many videos can be too small to extract meaningful spatial-temporal relations and achieve robust performance. To this end, we propose to decompose each video into a series of expression snippets, each of which contains a small number of facial movements, and attempt to augment the Transformer's ability for modeling intra-snippet and inter-snippet visual relations, respectively, obtaining the Expression snippet Transformer (EST). In particular, for intra-snippet modeling, we devise an attention-augmented snippet feature extractor (AA-SFE) to enhance the encoding of subtle facial movements of each snippet by gradually attending to more salient information. In addition, for inter-snippet modeling, we introduce a shuffled snippet order prediction (SSOP) head and a corresponding loss to improve the modeling of subtle motion changes across subsequent snippets by training the Transformer to identify shuffled snippet orders. Extensive experiments on four challenging datasets (i.e., BU-3DFE, MMI, AFEW, and DFEW) demonstrate that our EST is superior to other CNN-based methods, obtaining state-of-the-art performance.

cs.CV↗

arXiv2021

Deformation mechanisms of Mg-Ca-Zn alloys studied by means of micropillar compression tests

The effect of Ca and Zn in solid solution on the critical resolved shear stress (CRSS) of basal slip, tensile twinning and pyramidal slip in Mg alloys has been measured through compression tests on single crystal micropillars with different orientations. The solute atoms increased the CRSS for basal slip to ~ 13.5 MPa, while the CRSS for pyramidal slip was lower than 85 MPa, reducing significantly the plastic anisotropy in comparison with pure Mg. Moreover, the CRSSs for twin nucleation and growth were very similar (~ 37 MPa) and the large value of the CRSS for twin growth hindered the growth of twins during thermo-mechanical processing. Finally, evidence of prismatic slip and cross-slip between basal and prismatic dislocations was found. It is concluded that the reduction of plastic anisotropy, the activation of different slip systems and cross-slip and the weak basal texture promoted by the large CRSS for twin growth are responsible for the improved ductility and formability of Mg-Ca-Zn alloys.

cond-mat.mtrl-sci↗

arXiv2021

Intersections of Deligne--Lusztig varieties and Springer fibres

In this paper we prove a direct geometric relation between Deligne--Lusztig varieties and Springer fibres in type $\mathsf{A}$: For any rational unipotent element, the Springer fibre cuts out a unique component of a specific Deligne--Lusztig variety; moreover, this component forms an open dense subset of a component of the Springer fibre. This boils down to a map from the unipotent variety to the Weyl group, and combines several constructions with a combinatorial flavour (like Weyr normal forms, Robinson--Schensted correspondence, and Spaltenstein's and Steinberg's labellings); it also provides a geometric interpretation of a classical dimension formula of unipotent centralisers.

math.RT↗

arXiv2021

Towards High Performance Human Keypoint Detection

Human keypoint detection from a single image is very challenging due to occlusion, blur, illumination and scale variance. In this paper, we address this problem from three aspects by devising an efficient network structure, proposing three effective training strategies, and exploiting four useful postprocessing techniques. First, we find that context information plays an important role in reasoning human body configuration and invisible keypoints. Inspired by this, we propose a cascaded context mixer (CCM), which efficiently integrates spatial and channel context information and progressively refines them. Then, to maximize CCM's representation capability, we develop a hard-negative person detection mining strategy and a joint-training strategy by exploiting abundant unlabeled data. It enables CCM to learn discriminative features from massive diverse poses. Third, we present several sub-pixel refinement techniques for postprocessing keypoint predictions to improve detection accuracy. Extensive experiments on the MS COCO keypoint detection benchmark demonstrate the superiority of the proposed method over representative state-of-the-art (SOTA) methods. Our single model achieves comparable performance with the winner of the 2018 COCO Keypoint Detection Challenge. The final ensemble model sets a new SOTA on this benchmark.

cs.CV↗

arXiv2021

Dynamic Attention guided Multi-Trajectory Analysis for Single Object Tracking

Most of the existing single object trackers track the target in a unitary local search window, making them particularly vulnerable to challenging factors such as heavy occlusions and out-of-view movements. Despite the attempts to further incorporate global search, prevailing mechanisms that cooperate local and global search are relatively static, thus are still sub-optimal for improving tracking performance. By further studying the local and global search results, we raise a question: can we allow more dynamics for cooperating both results? In this paper, we propose to introduce more dynamics by devising a dynamic attention-guided multi-trajectory tracking strategy. In particular, we construct dynamic appearance model that contains multiple target templates, each of which provides its own attention for locating the target in the new frame. Guided by different attention, we maintain diversified tracking results for the target to build multi-trajectory tracking history, allowing more candidates to represent the true target trajectory. After spanning the whole sequence, we introduce a multi-trajectory selection network to find the best trajectory that delivers improved tracking performance. Extensive experimental results show that our proposed tracking strategy achieves compelling performance on various large-scale tracking benchmarks. The project page of this paper can be found at https://sites.google.com/view/mt-track/.

cs.CV↗

arXiv2021

Flatland Competition 2020: MAPF and MARL for Efficient Train Coordination on a Grid World

The Flatland competition aimed at finding novel approaches to solve the vehicle re-scheduling problem (VRSP). The VRSP is concerned with scheduling trips in traffic networks and the re-scheduling of vehicles when disruptions occur, for example the breakdown of a vehicle. While solving the VRSP in various settings has been an active area in operations research (OR) for decades, the ever-growing complexity of modern railway networks makes dynamic real-time scheduling of traffic virtually impossible. Recently, multi-agent reinforcement learning (MARL) has successfully tackled challenging tasks where many agents need to be coordinated, such as multiplayer video games. However, the coordination of hundreds of agents in a real-life setting like a railway network remains challenging and the Flatland environment used for the competition models these real-world properties in a simplified manner. Submissions had to bring as many trains (agents) to their target stations in as little time as possible. While the best submissions were in the OR category, participants found many promising MARL approaches. Using both centralized and decentralized learning based approaches, top submissions used graph representations of the environment to construct tree-based observations. Further, different coordination mechanisms were implemented, such as communication and prioritization between agents. This paper presents the competition setup, four outstanding solutions to the competition, and a cross-comparison between them.

cs.AI↗

arXiv2020

Sedimentation of a Colloidal Monolayer Down an Inclined Plane

We study the driven collective dynamics of a colloidal monolayer sedimentating down an inclined plane. The action of the gravity force parallel to the bottom wall creates a flow around each colloid, and the hydrodynamic interactions among the colloids accelerate the sedimentation as the local density increases. This leads to the creation of a universal "triangular" inhomogeneous density profile, with a traveling density shock at the leading front moving in the downhill direction. Unlike density shocks in a colloidal monolayer driven by applied torques rather than forces [Phys. Rev. Fluids, 2(9):092301, 2017], the density front during sedimentation remains stable over long periods of time even though it develops a roughness on the order of tens of particle diameters. Through experimental measurements and particle-based computer simulations, we find that the Burgers equation can model the density profile along the sedimentation direction as a function of time remarkably well, with a modest improvement if the nonlinear conservation law accounts for the sub-linear dependence of the collective sedimentation velocity on density.

cond-mat.soft↗

arXiv2020

Recent Progress in Appearance-based Action Recognition

Action recognition, which is formulated as a task to identify various human actions in a video, has attracted increasing interest from computer vision researchers due to its importance in various applications. Recently, appearance-based methods have achieved promising progress towards accurate action recognition. In general, these methods mainly fulfill the task by applying various schemes to model spatial and temporal visual information effectively. To better understand the current progress of appearance-based action recognition, we provide a comprehensive review of recent achievements in this area. In particular, we summarise and discuss several dozens of related research papers, which can be roughly divided into four categories according to different appearance modelling strategies. The obtained categories include 2D convolutional methods, 3D convolutional methods, motion representation-based methods, and context representation-based methods. We analyse and discuss representative methods from each category, comprehensively. Empirical results are also summarised to better illustrate cutting-edge algorithms. We conclude by identifying important areas for future research gleaned from our categorisation.

cs.CV↗

arXiv2020

CL-MAPF: Multi-Agent Path Finding for Car-Like Robots with Kinematic and Spatiotemporal Constraints

Multi-Agent Path Finding has been widely studied in the past few years due to its broad application in the field of robotics and AI. However, previous solvers rely on several simplifying assumptions. They limit their applicability in numerous real-world domains that adopt nonholonomic car-like agents rather than holonomic ones. In this paper, we give a mathematical formalization of Multi-Agent Path Finding for Car-Like robots (CL-MAPF) problem. For the first time, we propose a novel hierarchical search-based solver called Car-like Conflict-Based Search to address this problem. It applies a body conflict tree to address collisions considering shapes of the agents. We introduce a new algorithm called Spatiotemporal Hybrid-State A* as the single-agent path planner to generate path satisfying both kinematic and spatiotemporal constraints. We also present a sequential planning version of our method for the sake of efficiency. We compare our method with two baseline algorithms on a dedicated benchmark containing 3000 instances and validate it in real-world scenarios. The experiment results give clear evidence that our algorithm scales well to a large number of agents and is able to produce solutions that can be directly applied to car-like robots in the real world. The benchmark and source code are released in https://github.com/APRIL-ZJU/CL-CBS.

cs.RO↗

arXiv2020

Beyond Point Estimate: Inferring Ensemble Prediction Variation from Neuron Activation Strength in Recommender Systems

Despite deep neural network (DNN)'s impressive prediction performance in various domains, it is well known now that a set of DNN models trained with the same model specification and the same data can produce very different prediction results. Ensemble method is one state-of-the-art benchmark for prediction uncertainty estimation. However, ensembles are expensive to train and serve for web-scale traffic. In this paper, we seek to advance the understanding of prediction variation estimated by the ensemble method. Through empirical experiments on two widely used benchmark datasets MovieLens and Criteo in recommender systems, we observe that prediction variations come from various randomness sources, including training data shuffling, and parameter random initialization. By introducing more randomness into model training, we notice that ensemble's mean predictions tend to be more accurate while the prediction variations tend to be higher. Moreover, we propose to infer prediction variation from neuron activation strength and demonstrate the strong prediction power from activation strength features. Our experiment results show that the average R squared on MovieLens is as high as 0.56 and on Criteo is 0.81. Our method performs especially well when detecting the lowest and highest variation buckets, with 0.92 AUC and 0.89 AUC respectively. Our approach provides a simple way for prediction variation estimation, which opens up new opportunities for future work in many interesting areas (e.g.,model-based reinforcement learning) without relying on serving expensive ensemble models.

cs.LG↗

arXiv2020

Condensing Two-stage Detection with Automatic Object Key Part Discovery

Modern two-stage object detectors generally require excessively large models for their detection heads to achieve high accuracy. To address this problem, we propose that the model parameters of two-stage detection heads can be condensed and reduced by concentrating on object key parts. To this end, we first introduce an automatic object key part discovery task to make neural networks discover representative sub-parts in each foreground object. With these discovered key parts, we then decompose the object appearance modeling into a key part modeling process and a global modeling process for detection. Key part modeling encodes fine and detailed features from the discovered key parts, and global modeling encodes rough and holistic object characteristics. In practice, such decomposition allows us to significantly abridge model parameters without sacrificing much detection accuracy. Experiments on popular datasets illustrate that our proposed technique consistently maintains original performance while waiving around 50% of the model parameters of common two-stage detection heads, with the performance only deteriorating by 1.5% when waiving around 96% of the original model parameters. Codes are released on: https://github.com/zhechen/Condensing2stageDetection.

cs.CV↗

arXiv2020

Invertible Neural BRDF for Object Inverse Rendering

We introduce a novel neural network-based BRDF model and a Bayesian framework for object inverse rendering, i.e., joint estimation of reflectance and natural illumination from a single image of an object of known geometry. The BRDF is expressed with an invertible neural network, namely, normalizing flow, which provides the expressive power of a high-dimensional representation, computational simplicity of a compact analytical model, and physical plausibility of a real-world BRDF. We extract the latent space of real-world reflectance by conditioning this model, which directly results in a strong reflectance prior. We refer to this model as the invertible neural BRDF model (iBRDF). We also devise a deep illumination prior by leveraging the structural bias of deep neural networks. By integrating this novel BRDF model and reflectance and illumination priors in a MAP estimation formulation, we show that this joint estimation can be computed efficiently with stochastic gradient descent. We experimentally validate the accuracy of the invertible neural BRDF model on a large number of measured data and demonstrate its use in object inverse rendering on a number of synthetic and real images. The results show new ways in which deep neural networks can help solve challenging radiometric inverse problems.

cs.CV↗

arXiv2020

Block Shuffle: A Method for High-resolution Fast Style Transfer with Limited Memory

Fast Style Transfer is a series of Neural Style Transfer algorithms that use feed-forward neural networks to render input images. Because of the high dimension of the output layer, these networks require much memory for computation. Therefore, for high-resolution images, most mobile devices and personal computers cannot stylize them, which greatly limits the application scenarios of Fast Style Transfer. At present, the two existing solutions are purchasing more memory and using the feathering-based method, but the former requires additional cost, and the latter has poor image quality. To solve this problem, we propose a novel image synthesis method named \emph{block shuffle}, which converts a single task with high memory consumption to multiple subtasks with low memory consumption. This method can act as a plug-in for Fast Style Transfer without any modification to the network architecture. We use the most popular Fast Style Transfer repository on GitHub as the baseline. Experiments show that the quality of high-resolution images generated by our method is better than that of the feathering-based method. Although our method is an order of magnitude slower than the baseline, it can stylize high-resolution images with limited memory, which is impossible with the baseline. The code and models will be made available on \url{https://github.com/czczup/block-shuffle}.

cs.CV↗

arXiv2020

Fisher Discriminative Least Squares Regression for Image Classification

Discriminative least squares regression (DLSR) has been shown to achieve promising performance in multi-class image classification tasks. Its key idea is to force the regression labels of different classes to move in opposite directions by means of the proposed the joint use of the $ε$-draggings technique, yielding discriminative regression model exhibiting wider margins, and the Fisher criterion. The $ε$-draggings technique ignores an important problem: its non-negative relaxation matrix is dynamically updated in optimization, which means the dragging values can also cause the labels from the same class to be uncorrelated. In order to learn a more powerful discriminative projection, as well as regression labels, we propose a Fisher regularized DLSR (FDLSR) framework by constraining the relaxed labels using the Fisher criterion. On one hand, the Fisher criterion improves the intra-class compactness of the relaxed labels during relaxation learning. On the other hand, it is expected further to enhance the inter-class separability of $ε$-draggings technique. FDLSR for the first time ever attempts to integrate the Fisher discriminant criterion and $ε$-draggings technique into one unified model because they are absolutely complementary in learning discriminative projection. Extensive experiments on various datasets demonstrate that the proposed FDLSR method achieves performance that is superior to other state-of-the-art classification methods. The Matlab codes of this paper are available at https://github.com/chenzhe207/FDLSR.

cs.CV↗

arXiv2020

An Executable Operational Semantics for Rust with the Formalization of Ownership and Borrowing

Rust is an emergent systems programming language highlighting memory safety by its Ownership and Borrowing System (OBS). The existing formal semantics for Rust only covers limited subsets of the major language features of Rust. Moreover, they formalize OBS as type systems at the language-level, which can only be used to conservatively analyze programs against the OBS invariants at compile-time. That is, they are not executable, and thus cannot be used for automated verification of runtime behavior. In this paper, we propose RustSEM, a new executable operational semantics for Rust. RustSEM covers a much larger subset of the major language features than existing semantics. Moreover, RustSEM provides an operational semantics for OBS at the memory-level, which can be used to verify the runtime behavior of Rust programs against the OBS invariants. We have implemented RustSEM in the executable semantics modeling tool K-Framework. We have evaluated the semantics correctness of RustSEM wrt. the Rust compiler using around 700 tests. In particular, we have proposed a new technique for testing semantic consistency to ensure the absence of semantic ambiguities on all possible execution selections. We have also evaluated the potential applications of RustSEM in automated runtime and formal verification for both functional and memory properties. Experimental results show that RustSEM can enhance the memory safety mechanism of Rust, as it is more powerful than OBS in detecting memory errors.

cs.PL↗

arXiv2020

A note on cusp forms and representations of $\mathrm{SL}_2(\mathbb{F}_p)$

Cusp forms are certain holomorphic functions defined on the upper half-plane, and the space of cusp forms for the principal congruence subgroup $Γ(p)$, $p$ a prime, is acted by $\mathrm{SL}_2(\mathbb{F}_p)$. Meanwhile, there is a finite field incarnation of the upper half-plane, the Deligne--Lusztig (or Drinfeld) curve, whose cohomology space is also acted by $\mathrm{SL}_2(\mathbb{F}_p)$. In this note we compute the relation between these two spaces in the weight $2$ case.

math.RT↗