Diffusion for De-Occlusion: Accessory-Aware Diffusion Inpainting for Robust Ear Biometric Recognition
Image generated by Gemini AI
A study evaluates a diffusion-based ear inpainting technique aimed at improving ear recognition systems hindered by occlusions from accessories like earrings and earphones. The model reconstructs occluded ear regions while maintaining anatomical accuracy. Tests across various vision transformer models reveal that this technique enhances recognition performance, demonstrating its practical utility in biometric applications.
Diffusion Technique Enhances Ear Biometric Recognition Amid Occlusions
Recent research highlights a diffusion-based ear inpainting method designed to address performance challenges in ear biometric recognition systems caused by occlusions from accessories like earrings and earphones. This approach aims to improve the efficacy of transformer-based recognition systems in unconstrained imaging scenarios.
Ear occlusions significantly hinder the accuracy of biometric systems relying on ear morphology. The study introduces a novel pre-processing aid that utilizes a diffusion model to reconstruct occluded ear regions, effectively restoring key anatomical features such as the helix and lobule.
Methodology and Evaluation
The research tested the diffusion-based inpainting technique across various transformer models on benchmark datasets. An accessory mask is generated from the input ear image, guiding the inpainting process to focus on areas affected by occlusions.
Experiments revealed that the diffusion-based inpainting method significantly enhances recognition performance, with improved accuracy levels for systems implementing inpainting compared to those that did not.
Related Topics:
📰 Original Source: https://arxiv.org/abs/2601.19795v1
All rights and credit belong to the original publisher.