Deep Cascaded Bi-Network for Face Hallucination

TL;DR

This paper introduces a deep cascaded bi-network framework that jointly optimizes face hallucination and dense correspondence estimation to generate high-quality, detailed faces from very low-resolution images with unconstrained poses.

Contribution

It proposes a novel gated deep bi-network with two specialized branches and a unified optimization framework for face hallucination and dense correspondence estimation.

Findings

Achieves high-quality face hallucination on in-the-wild low-res images

Handles significant pose and illumination variations

Outperforms existing methods in hallucination quality

Abstract

We present a novel framework for hallucinating faces of unconstrained poses and with very low resolution (face size as small as 5pxIOD). In contrast to existing studies that mostly ignore or assume pre-aligned face spatial configuration (e.g. facial landmarks localization or dense correspondence field), we alternatingly optimize two complementary tasks, namely face hallucination and dense correspondence field estimation, in a unified framework. In addition, we propose a new gated deep bi-network that contains two functionality-specialized branches to recover different levels of texture details. Extensive experiments demonstrate that such formulation allows exceptional hallucination quality on in-the-wild low-res faces with significant pose and illumination variations.