Visual Question Answering in Remote Sensing with Cross-Attention and Multimodal Information Bottleneck | IEEE Conference Publication | IEEE Xplore