Video super-resolution via mixed spatial-temporal convolution and selective fusion

作者：

Highlights：

• A Mixed Spatial-Temporal Convolution block combines spatial-temporal information.

• Selective feature fusion strategy ensures the effectiveness of feature merging.

• Motion compensation alleviates the influence of misalignments.

• Our approach can achieve better performance on various datasets.

摘要

•A Mixed Spatial-Temporal Convolution block combines spatial-temporal information.•Selective feature fusion strategy ensures the effectiveness of feature merging.•Motion compensation alleviates the influence of misalignments.•Our approach can achieve better performance on various datasets.

论文关键词：Video super-Resolution,Mixed spatial-Temporal convolution,Selective feature fusion

论文评审过程：Received 20 October 2020, Revised 10 January 2022, Accepted 3 February 2022, Available online 5 February 2022, Version of Record 18 February 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.108577