FineNet: Frame Interpolation and Enhancement for Face Video Deblurring

TL;DR

FineNet is a novel face video deblurring method that combines frame enhancement and interpolation, utilizing face structure-aware modules to significantly outperform previous techniques in quality and robustness.

Contribution

The paper introduces a dual-approach framework for face video deblurring, integrating a face-structure-aware module to improve interpolation and enhancement results.

Findings

Outperforms previous state-of-the-art methods significantly

Effective on both real and synthetic blurry videos

Improves deblurring quality through combined enhancement and interpolation

Abstract

The objective of this work is to deblur face videos. We propose a method that tackles this problem from two directions: (1) enhancing the blurry frames, and (2) treating the blurry frames as missing values and estimate them by interpolation. These approaches are complementary to each other, and their combination outperforms individual ones. We also introduce a novel module that leverages the structure of faces for finding positional offsets between video frames. This module can be integrated into the processing pipelines of both approaches, improving the quality of the final outcome. Experiments on three real and synthetically generated blurry video datasets show that our method outperforms the previous state-of-the-art methods by a large margin in terms of both quantitative and qualitative results.