Deblur-Avatar: Animatable Avatars from Motion-Blurred Monocular Videos

TL;DR

This paper presents a new framework for creating high-quality, animatable 3D human avatars from motion-blurred monocular videos by explicitly modeling human motion trajectories during exposure, improving reconstruction and rendering quality.

Contribution

It introduces a human movement-based motion blur model integrated into 3D Gaussian Splatting, enabling sharp avatar reconstruction from motion-blurred videos, which was not addressed by prior methods.

Findings

Outperforms existing methods in rendering quality and metrics

Produces sharp, high-fidelity avatars from blurred videos

Enables real-time rendering under motion blur conditions

Abstract

We introduce a novel framework for modeling high-fidelity, animatable 3D human avatars from motion-blurred monocular video inputs. Motion blur is prevalent in real-world dynamic video capture, especially due to human movements in 3D human avatar modeling. Existing methods either (1) assume sharp image inputs, failing to address the detail loss introduced by motion blur, or (2) mainly consider blur by camera movements, neglecting the human motion blur which is more common in animatable avatars. Our proposed approach integrates a human movement-based motion blur model into 3D Gaussian Splatting (3DGS). By explicitly modeling human motion trajectories during exposure time, we jointly optimize the trajectories and 3D Gaussians to reconstruct sharp, high-quality human avatars. We employ a pose-dependent fusion mechanism to distinguish moving body regions, optimizing both blurred and sharp areas effectively. Extensive experiments on synthetic and real-world datasets demonstrate that our method significantly outperforms existing methods in rendering quality and quantitative metrics, producing sharp avatar reconstructions and enabling real-time rendering under challenging motion blur conditions.