Deep Compositional Captioning: Describing Novel Object Categories without Paired Training Data | IEEE Conference Publication | IEEE Xplore
Nothing Special   »   [go: up one dir, main page]