Time cost is a major challenge in achieving high-quality pluralistic image completion. Recently, the Retentive Network (RetNet) in natural language processing offers a novel approach to this problem with its low-cost inference capabilities. Inspired by this, we apply RetNet to the pluralistic image completion task in computer vision. We present RetCompletion, a two-stage framework. In the first stage, we introduce Bi-RetNet, a bidirectional sequence information fusion model that integrates contextual information from images. During inference, we employ a unidirectional pixel-wise update strategy to restore consistent image structures, achieving both high reconstruction quality and fast inference speed. In the second stage, we use a CNN for low-resolution upsampling to enhance texture details. Experiments on ImageNet and CelebA-HQ demonstrate that our inference speed is 10× faster than ICT and 15× faster than RePaint. The proposed RetCompletion significantly improves inference speed and delivers strong performance.
@article{arxiv.2410.04056,
title = {RetCompletion:High-Speed Inference Image Completion with Retentive Network},
author = {Yueyang Cang and Pingge Hu and Xiaoteng Zhang and Xingtong Wang and Yuhang Liu and Li Shi},
journal= {arXiv preprint arXiv:2410.04056},
year = {2024}
}