Spatial Cross-Attention for Transformer-Based Image Captioning.

Khoa Anh Ngo, Kyuhong Shim, Byonghyo Shim

Browse the full ICASSP paper archive.