Fast Vision Transformer via Additive Attention | IEEE Conference Publication | IEEE Xplore