Improving Goal-Oriented Visual Dialog Agents via Advanced Recurrent Nets with Tempered Policy Gradient

Open in new window