New ESTHER model reconstructs 3D hands from egocentric stereo vision

Researchers have introduced ESTHER, a novel end-to-end model for metric 3D hand reconstruction from egocentric stereo vision. This model is designed to handle the challenges of real-world capture, such as varying lighting, motion blur, and missing views. ESTHER achieves state-of-the-art accuracy and demonstrates superior generalization capabilities, even when reduced to a single monocular view, by leveraging stereo guidance to maintain metric scale. AI

IMPACT This research advances 3D hand reconstruction for robotics and AR/VR by providing a robust model and benchmark for egocentric stereo data.

RANK_REASON The cluster contains a research paper detailing a new model and dataset for 3D hand reconstruction. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• ESTHER
• ESTHER3D
• Hongyu Ma
• Hugging Face

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →