Improving Consistency and Correctness of Sequence Inpainting using Semantically Guided Generative Adversarial Network

Contemporary benchmark methods for image inpainting are based on deep generative models and specifically leverage adversarial loss for yielding realistic reconstructions. However, these models cannot be directly applied on image/video sequences because of an intrinsic drawbackthe reconstructions might be independently realistic, but, when visualized as a sequence, often lacks fidelity to the original uncorrupted sequence. The fundamental reason is that these methods try to find the best matching latent space representation near to natural image manifold without any explicit distance based loss. In this paper, we present a semantically conditioned Generative Adversarial Network (GAN) for sequence inpainting. The conditional information constrains the GAN to map a latent representation to a point in image manifold respecting the underlying pose and semantics of the scene. To the best of our knowledge, this is the first work which simultaneously addresses consistency and correctness of generative model based inpainting. We show that our generative model learns to disentangle pose and appearance information; this independence is exploited by our model to generate highly consistent reconstructions. The conditional information also aids the generator network in GAN to produce sharper images compared to the original GAN formulation. This helps in achieving more appealing inpainting performance. Though generic, our algorithm was targeted for inpainting on faces. When applied on CelebA and Youtube Faces datasets, the proposed method results in significant improvement over the current benchmark, both in terms of quantitative evaluation (Peak Signal to Noise Ratio) and human visual scoring over diversified combinations of resolutions and deformations. Figure 1. Exemplary success of our model in simultaneously preserving facial semantics(appearance and expressions) and improving inpaiting quality. Benchmark generative models such as DIP [49] are agnostic to holistic facial semantics and thus generate independently realistic, yet structurally inconsistent solutions.