Efficient masked feature and group attention network for stereo image super-resolution

Jianwen Song, Arcot Sowmya, Jien Kato, Changming Sun

Abstract: Highlights•An efficient masked feature and group attention network for stereo image super-resolution provides state-of-the-art performance at low computational costs.•An efficient masked feature block with enhanced capabilities of local feature extraction is designed to extract features efficiently.•A group-weighted cross-attention module and a group-weighted self-attention module are proposed to fully capture cross-view and intra-view information.