Visual Dialog

We introduce the task of Visual Dialog, which requires an AI agent to hold a meaningful dialog with humans in natural, conversational language about visual content. Specifically, given an image, a dialog history, and a question about the image, the agent has to ground the question in image, infer context from history, and answer the question accurately. Visual Dialog is disentangled enough from a specific downstream task so as to serve as a general test of machine intelligence, while being grounded in vision enough to allow objective evaluation of individual responses and benchmark progress. We develop a novel two-person chat data-collection protocol to curate a large-scale Visual Dialog dataset (VisDial). VisDial contains 1 dialog (10 question-answer pairs) on similar to 140k images from the COCO dataset, with a total of similar to 1.4M dialog question-answer pairs. We introduce a family of neural encoder-decoder models for Visual Dialog with 3 encoders (Late Fusion, Hierarchical Recurrent Encoder and Memory Network) and 2 decoders (generative and discriminative), which outperform a number of sophisticated baselines. We propose a retrieval-based evaluation protocol for Visual Dialog where the AI agent is asked to sort a set of candidate answers and evaluated on metrics such as mean-reciprocal-rank of human response. We quantify gap between machine and human performance on the Visual Dialog task via human studies. Our dataset, code, and trained models will be released publicly at visualdialog.org. Putting it all together, we demonstrate the first 'visual chatbot'!