arXiv subjects

Lixin Yang

Publications and source records attributed to Lixin Yang.

44 records · Page 3Linked to original sources

arXiv2021

Two-point functions from chiral kinetic theory in magnetized plasma

We study the two-point functions from chiral kinetic theory which characterize the response to perturbative vector and axial gauge fields in magnetized chiral plasma. In the lowest Landau level approximation, the solution of chiral kinetic equations gives density waves of electric and axial charges, which contain chiral magnetic wave implied by the axial anomaly and magnetic field. We then obtain the constitutive relations for covariant currents and stress tensor that involving the density waves. By considering the difference between consistent and covariant anomalies explicitly, the correlators of consistent currents and stress tensor satisfy derivative symmetry, and therefore allow an effective action for the perturbative gauge fields as the generating functional of the correlators. We also verify the derivative symmetry of the correlators agrees with the Onsager relations.

nucl-th↗

arXiv2021

HandTailor: Towards High-Precision Monocular 3D Hand Recovery

3D hand pose estimation and shape recovery are challenging tasks in computer vision. We introduce a novel framework HandTailor, which combines a learning-based hand module and an optimization-based tailor module to achieve high-precision hand mesh recovery from a monocular RGB image. The proposed hand module unifies perspective projection and weak perspective projection in a single network towards accuracy-oriented and in-the-wild scenarios. The proposed tailor module then utilizes the coarsely reconstructed mesh model provided by the hand module as initialization, and iteratively optimizes an energy function to obtain better results. The tailor module is time-efficient, costs only 8ms per frame on a modern CPU. We demonstrate that HandTailor can get state-of-the-art performance on several public benchmarks, with impressive qualitative results on in-the-wild experiments. Code and video are available on our project webpage https://sites.google.com/view/handtailor.

cs.CV↗

arXiv2021

CPF: Learning a Contact Potential Field to Model the Hand-Object Interaction

Modeling the hand-object (HO) interaction not only requires estimation of the HO pose, but also pays attention to the contact due to their interaction. Significant progress has been made in estimating hand and object separately with deep learning methods, simultaneous HO pose estimation and contact modeling has not yet been fully explored. In this paper, we present an explicit contact representation namely Contact Potential Field (CPF), and a learning-fitting hybrid framework namely MIHO to Modeling the Interaction of Hand and Object. In CPF, we treat each contacting HO vertex pair as a spring-mass system. Hence the whole system forms a potential field with minimal elastic energy at the grasp position. Extensive experiments on the two commonly used benchmarks have demonstrated that our method can achieve state-of-the-art in several reconstruction metrics, and allow us to produce more physically plausible HO pose even when the ground-truth exhibits severe interpenetration or disjointedness. Our code is available at https://github.com/lixiny/CPF.

cs.CV↗

arXiv2021

Magneto-vortical Effect in Strong Magnetic Field

We develop covariant chiral kinetic theory with Landau level basis. We use it to investigate a magnetized plasma with a transverse electric field and a steady vorticity as perturbations. After taking into account vacuum shift in the latter case, we find the resulting current and stress tensor in both cases can be matched consistently with constitutive equations of magnetohydrodynamics. We find the solution in the vorticity case contains both shifts in temperature and chemical potential as well as excitations of the lowest Landau level states. The solution gives rise to an vector charge density and axial current density. The vacuum parts coming from both shifts and excitations agree with previous studies and the medium parts coming entirely from excitations leads to a new contribution to vector charge and axial current density consistent with standard chiral vortical effect.

nucl-th↗

arXiv2020

Learning Universal Shape Dictionary for Realtime Instance Segmentation

We present a novel explicit shape representation for instance segmentation. Based on how to model the object shape, current instance segmentation systems can be divided into two categories, implicit and explicit models. The implicit methods, which represent the object mask/contour by intractable network parameters, and produce it through pixel-wise classification, are predominant. However, the explicit methods, which parameterize the shape with simple and explainable models, are less explored. Since the operations to generate the final shape are light-weighted, the explicit methods have a clear speed advantage over implicit methods, which is crucial for real-world applications. The proposed USD-Seg adopts a linear model, sparse coding with dictionary, for object shapes. First, it learns a dictionary from a large collection of shape datasets, making any shape being able to be decomposed into a linear combination through the dictionary. Hence the name "Universal Shape Dictionary". Then it adds a simple shape vector regression head to ordinary object detector, giving the detector segmentation ability with minimal overhead. For quantitative evaluation, we use both average precision (AP) and the proposed Efficiency of AP (AP$_E$) metric, which intends to also measure the computational consumption of the framework to cater to the requirements of real-world applications. We report experimental results on the challenging COCO dataset, in which our single model on a single Titan Xp GPU achieves 35.8 AP and 27.8 AP$_E$ at 65 fps with YOLOv4 as base detector, 34.1 AP and 28.6 AP$_E$ at 12 fps with FCOS as base detector.

cs.CV↗

arXiv2020

BiHand: Recovering Hand Mesh with Multi-stage Bisected Hourglass Networks

3D hand estimation has been a long-standing research topic in computer vision. A recent trend aims not only to estimate the 3D hand joint locations but also to recover the mesh model. However, achieving those goals from a single RGB image remains challenging. In this paper, we introduce an end-to-end learnable model, BiHand, which consists of three cascaded stages, namely 2D seeding stage, 3D lifting stage, and mesh generation stage. At the output of BiHand, the full hand mesh will be recovered using the joint rotations and shape parameters predicted from the network. Inside each stage, BiHand adopts a novel bisecting design which allows the networks to encapsulate two closely related information (e.g. 2D keypoints and silhouette in 2D seeding stage, 3D joints, and depth map in 3D lifting stage, joint rotations and shape parameters in the mesh generation stage) in a single forward pass. As the information represents different geometry or structure details, bisecting the data flow can facilitate optimization and increase robustness. For quantitative evaluation, we conduct experiments on two public benchmarks, namely the Rendered Hand Dataset (RHD) and the Stereo Hand Pose Tracking Benchmark (STB). Extensive experiments show that our model can achieve superior accuracy in comparison with state-of-the-art methods, and can produce appealing 3D hand meshes in several severe conditions.

cs.CV↗

arXiv2020

Chiral kinetic theory from Landau level basis

We derive a chiral kinetic theory with Landau level basis, which is valid for slow-varying magnetic field with arbitrary magnitude. We apply the new chiral kinetic theory to calculate the electric conductivity transverse to the magnetic field in a magnetized QED and QCD plasma. Under the lowest Landau level approximation and relaxation time approximation, we find the transverse conductivity approaches a constant in the large magnetic field limit and is inversely proportional to the relaxation time. We also obtain a frequency-dependent transverse conductivity in response to a time-dependent electric field. We find a high frequency enhancement in this conductivity.

nucl-th↗

arXiv2018

On Mass correction to Chiral Vortical Effect and Chiral Separation Effect

We study coefficients of axial chiral vortical effect and chiral separation effect at finite temperature and vector chemical potential in massive theories. We present two independent methods of calculating the coefficients: one from field theory and the other using the mass term in axial anomaly equation. An ambiguity in the integration constant similar to hydrodynamic approach to axial chiral vortical effect exists in the latter, but can be fixed naturally in the presence of mass. We obtain perfect agreement between the methods. The results of axial chiral vortical effect and chiral separation effect indicate that the presence of mass generically suppresses the two coefficients, with less suppression at larger chemical potential. For phenomenologically relevant case of quark gluon plasma with three quark flavor, we find the correction is negligible.

nucl-th↗