arXiv subjects

Jianqiao Hu

Publications and source records attributed to Jianqiao Hu.

5 recordsLinked to original sources

arXiv2026

A stochastic model for elastoplastic contact of rough surfaces incorporating scale-dependent hardness

The stress concentrations caused by inherent roughness of natural and manufactured surfaces often induce plastic deformation at contact interfaces, a challenge compounded by competing influences of the size effect of plastic deformation and self-affine rough surface topography. To address this, we developed a novel methodology based on stochastic theory using compounded Chapman-Kolmogorov equations, for the first time, to solve elastoplastic contact problems involving scale-dependent hardness. Our approach formulates three integral equations describing the evolution of probability density functions of elastic contact pressure, relative plastic contact area, and relative non-contact area across geometrical scales. We thoroughly investigate the effects of scale-dependent hardness on contact pressure distribution, relative elastic and plastic contact areas, and the area-to-load relationship. By adjusting various mechanical and material properties, our model predicts a smooth transition from linear elasticity to elastic-plastic behavior and finally to full plasticity. A key advancement is the derivation of a new topographic yield parameter incorporating a wider range of material and geometrical properties, aiding identification of contact status. Numerical solutions enable highly precise determination of elastic and plastic limits via curve-fitting, and we also provide a new diagram for rapid identification of contact status. This study pioneers a stochastic process framework for applying the compounded Chapman-Kolmogorov equation to rough surface contact analysis, and the integral equations characterizing how interfacial properties evolve with scale could offer valuable insights for other multidisciplinary fields where multiscale roughness is critical, such as earthquakes, electrical contact, and contact electrification.

cond-mat.stat-mech↗

arXiv2025

M3DHMR: Monocular 3D Hand Mesh Recovery

Monocular 3D hand mesh recovery is challenging due to high degrees of freedom of hands, 2D-to-3D ambiguity and self-occlusion. Most existing methods are either inefficient or less straightforward for predicting the position of 3D mesh vertices. Thus, we propose a new pipeline called Monocular 3D Hand Mesh Recovery (M3DHMR) to directly estimate the positions of hand mesh vertices. M3DHMR provides 2D cues for 3D tasks from a single image and uses a new spiral decoder consist of several Dynamic Spiral Convolution (DSC) Layers and a Region of Interest (ROI) Layer. On the one hand, DSC Layers adaptively adjust the weights based on the vertex positions and extract the vertex features in both spatial and channel dimensions. On the other hand, ROI Layer utilizes the physical information and refines mesh vertices in each predefined hand region separately. Extensive experiments on popular dataset FreiHAND demonstrate that M3DHMR significantly outperforms state-of-the-art real-time methods.

cs.CV↗

arXiv2025

EmoFace: Emotion-Content Disentangled Speech-Driven 3D Talking Face Animation

The creation of increasingly vivid 3D talking face has become a hot topic in recent years. Currently, most speech-driven works focus on lip synchronisation but neglect to effectively capture the correlations between emotions and facial motions. To address this problem, we propose a two-stream network called EmoFace, which consists of an emotion branch and a content branch. EmoFace employs a novel Mesh Attention mechanism to analyse and fuse the emotion features and content features. Particularly, a newly designed spatio-temporal graph-based convolution, SpiralConv3D, is used in Mesh Attention to learn potential temporal and spatial feature dependencies between mesh vertices. In addition, to the best of our knowledge, it is the first time to introduce a new self-growing training scheme with intermediate supervision to dynamically adjust the ratio of groundtruth adopted in the 3D face animation task. Comprehensive quantitative and qualitative evaluations on our high-quality 3D emotional facial animation dataset, 3D-RAVDESS ($4.8863\times 10^{-5}$mm for LVE and $0.9509\times 10^{-5}$mm for EVE), together with the public dataset VOCASET ($2.8669\times 10^{-5}$mm for LVE and $0.4664\times 10^{-5}$mm for EVE), demonstrate that our approach achieves state-of-the-art performance.

cs.CV↗

arXiv2023

MVHumanNet: A Large-scale Dataset of Multi-view Daily Dressing Human Captures

In this era, the success of large language models and text-to-image models can be attributed to the driving force of large-scale datasets. However, in the realm of 3D vision, while remarkable progress has been made with models trained on large-scale synthetic and real-captured object data like Objaverse and MVImgNet, a similar level of progress has not been observed in the domain of human-centric tasks partially due to the lack of a large-scale human dataset. Existing datasets of high-fidelity 3D human capture continue to be mid-sized due to the significant challenges in acquiring large-scale high-quality 3D human data. To bridge this gap, we present MVHumanNet, a dataset that comprises multi-view human action sequences of 4,500 human identities. The primary focus of our work is on collecting human data that features a large number of diverse identities and everyday clothing using a multi-view human capture system, which facilitates easily scalable data collection. Our dataset contains 9,000 daily outfits, 60,000 motion sequences and 645 million frames with extensive annotations, including human masks, camera parameters, 2D and 3D keypoints, SMPL/SMPLX parameters, and corresponding textual descriptions. To explore the potential of MVHumanNet in various 2D and 3D visual tasks, we conducted pilot studies on view-consistent action recognition, human NeRF reconstruction, text-driven view-unconstrained human image generation, as well as 2D view-unconstrained human image and 3D avatar generation. Extensive experiments demonstrate the performance improvements and effective applications enabled by the scale provided by MVHumanNet. As the current largest-scale 3D human dataset, we hope that the release of MVHumanNet data with annotations will foster further innovations in the domain of 3D human-centric tasks at scale.

cs.CV↗

arXiv2020

Predicting the flow stress and dominant yielding mechanisms: analytical models based on discrete dislocation plasticity

Dislocations are the carriers of plasticity in crystalline materials. Their collective interaction behavior is dependent on the strain rate and sample size. In small specimens, details of the nucleation process are of particular importance. In the present work, discrete dislocation dynamics (DDD) simulations are performed to investigate the dominant yielding mechanisms in single crystalline copper pillars with diameters ranging from 100 to 800 nm. Based on our simulations with different strain rates and sample size, we observe a transition of the relevant nucleation mechanism from "dislocation multiplication" to "surface nucleation". Two physics-based analytical models are established to quantitatively predict this transition, showing a good agreement for different strain rates with our DDD simulation data and with available experimental data. Therefore, the proposed analytical models help to understand the interplay between different physical parameters and nucleation mechanisms and are well suitable to estimate the material strength for different material properties and under given loading conditions.

cond-mat.mtrl-sci↗