arXiv subjects

Zeyang Xia

Publications and source records attributed to Zeyang Xia.

3 recordsLinked to original sources

arXiv2025

MatchAttention: Embedding Explicit Matching Constraints into Attention for Efficient Stereo Matching

Standard attention mechanisms are not well suited to stereo matching. Global attention scales quadratically and provides no explicit matching constraint, while local attention is efficient but loses long-range correspondences. We propose MatchAttention, an attention mechanism that embeds an explicit matching constraint into attention by treating the relative position between a query and its matched key as a learnable component of attention sampling. Centering a small contiguous sampling window on this learnable relative position enforces the matching constraint and supports long-range correspondence at strictly linear attention complexity. A differentiable contiguous attention sampling (CAS) operator enables sub-pixel accuracy, and cascaded MatchAttention blocks iteratively refine the relative positions through residual connections. We instantiate MatchAttention as a hierarchical coarse-to-fine stereo network with two variants. MatchAttentionXL targets accuracy and MatchAttentionRT targets real-time edge inference. MatchAttentionXL achieves state-of-the-art accuracy on Middlebury V3 and top results across KITTI 2012/2015 and ETH3D. MatchAttentionRT runs at 9.3 ms on RTX 4060 Ti and 79.1 ms on Jetson Orin NX 16 GB at 1024 x 512, making it the first stereo model to deliver real-time edge inference without sacrificing zero-shot generalization. The code is available at https://github.com/TingmanYan/MatchAttention.

cs.CV↗

arXiv2025

Computational Orthodontic Force Simulation: A Review

In orthodontic treatment, the biological response of the tooth, periodontal ligament, and bone complex to orthodontic force is crucial in influencing treatment outcomes. The challenge lies in accurately measuring, estimating, and predicting these forces during clinical procedures. This review aims to fill the gap in the literature by systematically summarizing existing research on orthodontic force simulation, examining common loading techniques and technologies, and discussing the potential for refining the orthodontic force simulation process. The literature was comprehensively reviewed, with an emphasis on the exploration of the biological mechanism of tooth movement. Studies were categorized based on force-loading techniques for both fixed and invisible orthodontic appliances. Finite element (FE) analysis stands out as the predominant technique for orthodontic force simulation, with a significant focus on fixed orthodontics but limited emphasis on invisible orthodontics. Current orthodontic force simulations tend to be fragmented, often considering only the instantaneous response to applied forces. There exists an urgent demand for a sophisticated analytical simulation model. Such a model, possibly leveraging advanced technologies like deep learning, holds the promise of forecasting orthodontic treatment outcomes with heightened precision and efficiency.

physics.med-ph↗

arXiv2025

Safe Navigation for Robotic Digestive Endoscopy via Human Intervention-based Reinforcement Learning

With the increasing application of automated robotic digestive endoscopy (RDE), ensuring safe and efficient navigation in the unstructured and narrow digestive tract has become a critical challenge. Existing automated reinforcement learning navigation algorithms often result in potentially risky collisions due to the absence of essential human intervention, which significantly limits the safety and effectiveness of RDE in actual clinical practice. To address this limitation, we proposed a Human Intervention (HI)-based Proximal Policy Optimization (PPO) framework, dubbed HI-PPO, which incorporates expert knowledge to enhance RDE's safety. Specifically, HI-PPO combines Enhanced Exploration Mechanism (EEM), Reward-Penalty Adjustment (RPA), and Behavior Cloning Similarity (BCS) to address PPO's exploration inefficiencies for safe navigation in complex gastrointestinal environments. Comparative experiments were conducted on a simulation platform, and the results showed that HI-PPO achieved a mean ATE (Average Trajectory Error) of \(8.02\ \text{mm}\) and a Security Score of \(0.862\), demonstrating performance comparable to human experts. The code will be publicly available once this paper is published.

cs.RO↗