Blind Face Restoration

24 papers with code • 4 benchmarks • 4 datasets

Blind face restoration aims at recovering high-quality faces from the low-quality counterparts suffering from unknown degradation, such as low-resolution, noise, blur, compression artifacts, etc. When applied to real-world scenarios, it becomes more challenging, due to more complicated degradation, diverse poses and expressions.

Description source: Towards Real-World Blind Face Restoration with Generative Facial Prior

Image source: Towards Real-World Blind Face Restoration with Generative Facial Prior


Use these libraries to find Blind Face Restoration models and implementations
2 papers

Most implemented papers

DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better

KupynOrest/DeblurGANv2 ICCV 2019

We present a new end-to-end generative adversarial network (GAN) for single image motion deblurring, named DeblurGAN-v2, which considerably boosts state-of-the-art deblurring efficiency, quality, and flexibility.

HiFaceGAN: Face Renovation via Collaborative Suppression and Replenishment

Lotayou/Face-Renovation 11 May 2020

Existing face restoration researches typically relies on either the degradation prior or explicit guidance labels for training, which often results in limited generalization ability over real-world images with heterogeneous degradations and rich background contents.

GAN Prior Embedded Network for Blind Face Restoration in the Wild

yangxy/GPEN CVPR 2021

The proposed GAN prior embedded network (GPEN) is easy-to-implement, and it can generate visually photo-realistic results.

Blind Face Restoration: Benchmark Datasets and a Baseline Model

bitzpy/blind-face-restoration-benchmark-datasets-and-a-baseline-model 8 Jun 2022

To address this problem, we first synthesize two blind face restoration benchmark datasets called EDFace-Celeb-1M (BFR128) and EDFace-Celeb-150K (BFR512).

DifFace: Blind Face Restoration with Diffused Error Contraction

zsyoaoa/difface 13 Dec 2022

Moreover, the transition distribution can contract the error of the restoration backbone and thus makes our method more robust to unknown degradations.

Learning Warped Guidance for Blind Face Restoration

csxmli2016/GFRNet ECCV 2018

For better recovery of fine facial details, we modify the problem setting by taking both the degraded observation and a high-quality guided image of the same identity as input to our guided face restoration network (GFRNet).

Image Processing Using Multi-Code GAN Prior

genforce/mganprior CVPR 2020

Such an over-parameterization of the latent space significantly improves the image reconstruction quality, outperforming existing competitors.

Enhanced Blind Face Restoration With Multi-Exemplar Images and Adaptive Spatial Feature Fusion

csxmli2016/ASFFNet CVPR 2020

First, given a degraded observation, we select the optimal guidance based on the weighted affine distance on landmark sets, where the landmark weights are learned to make the guidance image optimized to HQ image reconstruction.

Blind Face Restoration via Deep Multi-scale Component Dictionaries

csxmli2016/DFDNet ECCV 2020

Next, with the degraded input, we match and select the most similar component features from their corresponding dictionaries and transfer the high-quality details to the input via the proposed dictionary feature transfer (DFT) block.

Progressive Semantic-Aware Style Transformation for Blind Face Restoration

chaofengc/PSFRGAN CVPR 2021

Compared with previous networks, the proposed PSFR-GAN makes full use of the semantic (parsing maps) and pixel (LQ images) space information from different scales of input pairs.