Analytic experts supply demonstrations for imitation learning of a goal-oriented visual dialog questioner, which is then refined via reinforcement learning to reach state-of-the-art on the GuessWhat?! benchmark.
VQA: Visual Question Answering,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning Goal-Oriented Visual Dialog Agents: Imitating and Surpassing Analytic Experts
Analytic experts supply demonstrations for imitation learning of a goal-oriented visual dialog questioner, which is then refined via reinforcement learning to reach state-of-the-art on the GuessWhat?! benchmark.