Learning Distinct and Representative Modes for Image Captioning

Open in new window