Learning to Deblur and Rotate Motion-Blurred Faces

TL;DR

This paper introduces a neural network approach to generate sharp, multi-view videos of faces from a single motion-blurred image, leveraging large datasets and a new multi-view face dataset for training.

Contribution

It presents a novel method that reconstructs 3D face videos from a single blurred image using multi-view constraints and a new dataset, improving face deblurring and view synthesis.

Findings

Successfully generates sharp multi-view videos from blurred images.

Outperforms existing methods in face deblurring and view synthesis.

Demonstrates generalization across diverse faces and expressions.

Abstract

We propose a solution to the novel task of rendering sharp videos from new viewpoints from a single motion-blurred image of a face. Our method handles the complexity of face blur by implicitly learning the geometry and motion of faces through the joint training on three large datasets: FFHQ and 300VW, which are publicly available, and a new Bern Multi-View Face Dataset (BMFD) that we built. The first two datasets provide a large variety of faces and allow our model to generalize better. BMFD instead allows us to introduce multi-view constraints, which are crucial to synthesizing sharp videos from a new camera view. It consists of high frame rate synchronized videos from multiple views of several subjects displaying a wide range of facial expressions. We use the high frame rate videos to simulate realistic motion blur through averaging. Thanks to this dataset, we train a neural network to reconstruct a 3D video representation from a single image and the corresponding face gaze. We then provide a camera viewpoint relative to the estimated gaze and the blurry image as input to an encoder-decoder network to generate a video of sharp frames with a novel camera viewpoint. We demonstrate our approach on test subjects of our multi-view dataset and VIDTIMIT.