English

DVD: A Diagnostic Dataset for Multi-step Reasoning in Video Grounded Dialogue

Artificial Intelligence 2021-06-15 v2 Computation and Language Machine Learning

Abstract

A video-grounded dialogue system is required to understand both dialogue, which contains semantic dependencies from turn to turn, and video, which contains visual cues of spatial and temporal scene variations. Building such dialogue systems is a challenging problem, involving various reasoning types on both visual and language inputs. Existing benchmarks do not have enough annotations to thoroughly analyze dialogue systems and understand their capabilities and limitations in isolation. These benchmarks are also not explicitly designed to minimise biases that models can exploit without actual reasoning. To address these limitations, in this paper, we present DVD, a Diagnostic Dataset for Video-grounded Dialogues. The dataset is designed to contain minimal biases and has detailed annotations for the different types of reasoning over the spatio-temporal space of video. Dialogues are synthesized over multiple question turns, each of which is injected with a set of cross-turn semantic relationships. We use DVD to analyze existing approaches, providing interesting insights into their abilities and limitations. In total, DVD is built from 11k11k CATER synthetic videos and contains 1010 instances of 1010-round dialogues for each video, resulting in more than 100k100k dialogues and 1M1M question-answer pairs. Our code and dataset are publicly available at https://github.com/facebookresearch/DVDialogues.

Keywords

Cite

@article{arxiv.2101.00151,
  title  = {DVD: A Diagnostic Dataset for Multi-step Reasoning in Video Grounded Dialogue},
  author = {Hung Le and Chinnadhurai Sankar and Seungwhan Moon and Ahmad Beirami and Alborz Geramifard and Satwik Kottur},
  journal= {arXiv preprint arXiv:2101.00151},
  year   = {2021}
}

Comments

20 pages, 14 figures, 8 tables

R2 v1 2026-06-23T21:40:47.128Z