A Unified Framework

Self-occlusion is common when capturing people in the wild, where the performers do not follow predefined motion scripts. This challenges existing monocular human reconstruction systems that assume full body visibility. We introduce Self-Occluded Avatar Recovery (SOAR), a method for complete human reconstruction from partial observations where parts of the body are entirely unobserved. SOAR leverages structural normal prior and generative diffusion prior to address such an ill-posed reconstruction problem. For structural normal prior, we model human with an reposable surfel model with well-defined and easily readable shapes. For generative diffusion prior, we perform an initial reconstruction and refine it using score distillation. On various benchmarks, we show that SOAR performs favorably than state-of-the-art reconstruction and generation methods, and on-par comparing to concurrent works.

Our Results

Novel View Rendering

We show novel view rendering results for our reconstructed avatars, including RGB, normal, and occlusion.



Animation

We animate our reconstructed avatars in novel poses.



Comparison

We compare our method with GART [1] and GaussianAvatar [2], showing the rendered RGB and normal maps for each. Our approach produces more realistic and detailed results in terms of both texture and structure.