Answer-Driven Visual State Estimator for Goal-Oriented Visual Dialogue

Abstract

A goal-oriented visual dialogue involves multi-turn interactions between two agents, Questioner and Oracle. During which, the answer given by Oracle is of great significance, as it provides golden response to what Questioner concerns. Based on the answer, Questioner updates its belief on target visual content and further raises another question. Notably, different answers drive into different visual beliefs and future questions. However, existing methods always indiscriminately encode answers after much longer questions, resulting in a weak utilization of answers. In this paper, we propose an Answer-Driven Visual State Estimator (ADVSE) to impose the effects of different answers on visual states. First, we propose an Answer-Driven Focusing Attention (ADFA) to capture the answerdriven effect on visual attention by sharpening question-related attention and adjusting it by answer-based logical operation at each turn. Then based on the focusing attention, we get the visual state estimation by Conditional Visual Information Fusion (CVIF), where overall information and difference information are fused conditioning on the question-answer state. We evaluate the proposed ADVSE to both question generator and guesser tasks on the large-scale GuessWhat?! dataset and achieve the state-of-the-art performances on both tasks. The qualitative results indicate that the ADVSE boosts the agent to generate highly efficient questions and obtains reliable visual attentions during the reasonable question generation and guess processes. CCS CONCEPTS • Computing methodologies → Computer vision tasks; Discourse, dialogue and pragmatics; Natural language generation; Computer vision representations.

Modeling Explicit Concerning States for Reinforcement Learning in Visual Dialogue

“…Works [1,2,23,30,34] that studies using RL to improve goal-oriented VD in the Guesswhat?! task setting [9] are also related.…”