Work

Identity-Preserving Diffusion for Face Restoration

Year: 2025

Type: conference-paper

Abstract: Face restoration is a critical task in computer vision, aiming to restore high-quality facial images from degraded inputs. In existing diffusion models, identity information is not well preserved when confronted with severely degradation. To address this challenge, we propose a Local Patch-Based Identity-Preserving Diffusion (LPIP-Diff) framework. Our local patch-based strategy leverages the interrelationships between neighboring patches to model highly structured facial context, which facilitates the restoration of fine-grained details and the preservation of identity-related features. We also introduce a fusion degradation estimation method that makes each overlapping area restored multiple times by adjacent patches, effectively restoring local details. The experimental results of LPIP-Diff on three publicly available datasets, including one severely degraded dataset, consistently demonstrate its superiority over the state-of-the-art methods in terms of both quantitative and qualitative evaluations, strikes a good balance between realism and fidelity, and enhances robustness against degradation.

Source: IEEE International Conference on Acoustics Speech and Signal Processing

DOI: 10.1109/icassp49660.2025.10888736

Authors: Xiaying Bai, Yuhao Yang, Wenming Yang, Rui Zhu, Jing-Hao Xue

Institutions: Tsinghua Shenzhen International Graduate School, Tsinghua University, University of London, City, University of London, University College London

Language: English

Topic: Advanced Image Processing Techniques

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funder: National Natural Science Foundation of China

Awards: G3370211328, G3523047508, G4653242361