Semi-supervised multimodal coreference resolution in image narrations

Open in new window