AvatarPose: Avatar-guided 3D Pose Estimation of Close Human Interaction from Sparse Multi-view Videos
This addresses a challenging issue in human motion capture for applications like animation or VR, but it is an incremental improvement over existing multi-view methods.
The paper tackles the problem of estimating 3D pose and shape for multiple closely interacting people from sparse multi-view videos, achieving state-of-the-art performance by using personalized implicit neural avatars as a prior to bypass noisy 2D detections and handle occlusions.
Despite progress in human motion capture, existing multi-view methods often face challenges in estimating the 3D pose and shape of multiple closely interacting people. This difficulty arises from reliance on accurate 2D joint estimations, which are hard to obtain due to occlusions and body contact when people are in close interaction. To address this, we propose a novel method leveraging the personalized implicit neural avatar of each individual as a prior, which significantly improves the robustness and precision of this challenging pose estimation task. Concretely, the avatars are efficiently reconstructed via layered volume rendering from sparse multi-view videos. The reconstructed avatar prior allows for the direct optimization of 3D poses based on color and silhouette rendering loss, bypassing the issues associated with noisy 2D detections. To handle interpenetration, we propose a collision loss on the overlapping shape regions of avatars to add penetration constraints. Moreover, both 3D poses and avatars are optimized in an alternating manner. Our experimental results demonstrate state-of-the-art performance on several public datasets.