Detail-revealing Deep Video Super-resolution
2017 IEEE International Conference on Computer Vision (ICCV)(2017)
摘要
Previous CNN-based video super-resolution approaches need to align multiple frames to the reference. In this paper, we show that proper frame alignment and motion compensation is crucial for achieving high quality results. We accordingly propose a `sub-pixel motion compensation' (SPMC) layer in a CNN framework. Analysis and experiments show the suitability of this layer in video SR. The final end-to-end, scalable CNN framework effectively incorporates the SPMC layer and fuses multiple frames to reveal image details. Our implementation can generate visually and quantitatively high-quality results, superior to current state-of-the-arts, without the need of parameter tuning.
更多查看译文
关键词
proper frame alignment,sub-pixel motion compensation layer,video SR,scalable CNN framework,SPMC layer,detail-revealing deep video super-resolution,CNN-based video super-resolution approaches,multiple frame alignment,multiple frame fusion
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络