HumanNeRF: Free-viewpoint Rendering of Moving People from Monocular Video

TL;DR

HumanNeRF enables photorealistic free-viewpoint rendering of moving humans from monocular videos by combining volumetric representation and learned motion fields, allowing arbitrary viewpoint synthesis at any video frame.

Contribution

The paper introduces HumanNeRF, a novel method that synthesizes free-viewpoint renderings of humans from monocular videos using a canonical volumetric model and learned motion fields.

Findings

Significant improvements over prior methods in free-viewpoint rendering quality.

Capable of rendering from arbitrary viewpoints and full 360-degree paths.

Effective in uncontrolled, real-world video scenarios.

Abstract

We introduce a free-viewpoint rendering method -- HumanNeRF -- that works on a given monocular video of a human performing complex body motions, e.g. a video from YouTube. Our method enables pausing the video at any frame and rendering the subject from arbitrary new camera viewpoints or even a full 360-degree camera path for that particular frame and body pose. This task is particularly challenging, as it requires synthesizing photorealistic details of the body, as seen from various camera angles that may not exist in the input video, as well as synthesizing fine details such as cloth folds and facial appearance. Our method optimizes for a volumetric representation of the person in a canonical T-pose, in concert with a motion field that maps the estimated canonical representation to every frame of the video via backward warps. The motion field is decomposed into skeletal rigid and non-rigid motions, produced by deep networks. We show significant performance improvements over prior work, and compelling examples of free-viewpoint renderings from monocular video of moving humans in challenging uncontrolled capture scenarios.