Visual Dialog: Do we really need it?

ACL 2020

Publication date: July 5, 2020

Shubham Agarwal, Trung Bui, Joon-Young Lee, Ioannis Konstas, Verena Rieser

Visual Dialog involves understanding the dialog history (what has been discussed previously) and the current question (what is asked), in addition to grounding information in the image, to generate the correct response. In this paper, we show that co-attention models that explicitly encode dialog history outperform models that don’t, achieving state-of-the-art performance (72 % NDCG on val set). However, we also expose the shortcomings of the crowd-sourcing dataset collection procedure by showing that history is indeed only required for a small amount of the data and that the current evaluation metric encourages generic replies. To that end, we propose a challenging subset (VisDialConv) of the VisDial val set and provide a benchmark of 63% NDCG.

Learn More