SiamTrans: Zero-Shot Multi-Frame Image Restoration with Pre-Trained Siamese Transformers
Lin Liu, Shanxin Yuan, Jianzhuang Liu, Xin Guo, Youliang Yan, Qi Tian
[AAAI-22] Main Track
Abstract:
We propose a novel zero-shot multi-frame image restoration method for removing unwanted obstruction elements (such as rains, snow, and moire patterns) that vary in successive frames.
It has three stages: transformer pre-training, zero-shot restoration, and hard patch refinement.
Using the pre-trained transformers, our model is able to tell the motion difference between the true image information and the obstructing elements.
For zero-shot image restoration, we design a novel model, termed SiamTrans, which is constructed by Siamese transformers, encoders, and decoders.
Each transformer has a temporal attention layer and several self-attention layers, to capture both temporal and spatial information of multiple frames.
Only unsupervisedly pre-trained on the denoising task, SiamTrans is tested on three different low-level vision tasks (deraining, demoireing, and desnowing).
Compared with related methods, SiamTrans achieves the best performances, even outperforming those with supervised learning.
It has three stages: transformer pre-training, zero-shot restoration, and hard patch refinement.
Using the pre-trained transformers, our model is able to tell the motion difference between the true image information and the obstructing elements.
For zero-shot image restoration, we design a novel model, termed SiamTrans, which is constructed by Siamese transformers, encoders, and decoders.
Each transformer has a temporal attention layer and several self-attention layers, to capture both temporal and spatial information of multiple frames.
Only unsupervisedly pre-trained on the denoising task, SiamTrans is tested on three different low-level vision tasks (deraining, demoireing, and desnowing).
Compared with related methods, SiamTrans achieves the best performances, even outperforming those with supervised learning.
Introduction Video
Sessions where this paper appears
-
Poster Session 4
Fri, February 25 5:00 PM - 6:45 PM (+00:00)
Red 3
-
Poster Session 8
Sun, February 27 12:45 AM - 2:30 AM (+00:00)
Red 3
-
Oral Session 8
Sun, February 27 2:30 AM - 3:45 AM (+00:00)
Red 3