Faster Unsupervised Semantic Inpainting: A GAN Based Approach
This work addresses faster and higher-quality inpainting for image and video processing applications, but it is incremental as it builds on existing GAN frameworks.
The paper tackled the problem of slow inference speed and visual quality in GAN-based unsupervised semantic inpainting by improving initialization in iterative optimization, achieving about 4.5-5x speedup for images and 80x for videos compared to baseline while enhancing reconstruction quality.
In this paper, we propose to improve the inference speed and visual quality of contemporary baseline of Generative Adversarial Networks (GAN) based unsupervised semantic inpainting. This is made possible with better initialization of the core iterative optimization involved in the framework. To our best knowledge, this is also the first attempt of GAN based video inpainting with consideration to temporal cues. On single image inpainting, we achieve about 4.5-5$\times$ speedup and 80$\times$ on videos compared to baseline. Simultaneously, our method has better spatial and temporal reconstruction qualities as found on three image and one video dataset.