Texture Memory-Augmented Deep Patch-Based Image Inpainting
This addresses the problem of generating realistic textures in image inpainting for computer vision applications, representing an incremental improvement by hybridizing existing approaches.
The paper tackled image inpainting by combining patch-based methods and deep networks to restore missing regions with high-quality texture and sharp details, achieving superior performance on Places, CelebA-HQ, and Paris Street-View datasets.
Patch-based methods and deep networks have been employed to tackle image inpainting problem, with their own strengths and weaknesses. Patch-based methods are capable of restoring a missing region with high-quality texture through searching nearest neighbor patches from the unmasked regions. However, these methods bring problematic contents when recovering large missing regions. Deep networks, on the other hand, show promising results in completing large regions. Nonetheless, the results often lack faithful and sharp details that resemble the surrounding area. By bringing together the best of both paradigms, we propose a new deep inpainting framework where texture generation is guided by a texture memory of patch samples extracted from unmasked regions. The framework has a novel design that allows texture memory retrieval to be trained end-to-end with the deep inpainting network. In addition, we introduce a patch distribution loss to encourage high-quality patch synthesis. The proposed method shows superior performance both qualitatively and quantitatively on three challenging image benchmarks, i.e., Places, CelebA-HQ, and Paris Street-View datasets.