Evaluating Visual Conversational Agents via Cooperative Human-AI Games

Open in new window