High-Quality Facial Feature Occlusion Repair Based on S-GANs

In the image restoration task, the excellent performance of the generative adversarial network (GAN) is impressive, but in the task of generative face region inpainting, there are still great challenges. Traditional model approaches are not very effective in maintaining global consistency among facial components and recovering fine facial details. To address this challenge, this paper proposes a facial restoration generation network combining a precise segmentation network and GAN to accurately detect the missing feature parts of the face and perform effective and fine-grained restoration generation. We validate the proposed model using different image quality evaluation methods and using several open-source face datasets, and experimentally demonstrate that our model outperforms other current state-of-the-art network models in terms of generated image quality and coherent naturalness of face features in face image restoration generation tasks.