An Image Captioning Model Based on SE-ResNest and EMSA | IEEE Conference Publication | IEEE Xplore