Visual Dialogue without Vision or Dialogue

Daniela Massiceti, Puneet K. Dokania, N. Siddharth, Philip Torr

Research output: Contribution to conferencePaperpeer-review

Abstract

We characterise some of the quirks and shortcomings in the exploration of visual dialogue (VD)—a sequential question-answering task where the questions and corresponding answers are related through given visual stimuli. To do so, we develop an embarrassingly simple method based on canonical correlation analysis (CCA) that, on the standard dataset, achieves near state-of-the-art performance on mean rank (MR). In direct contrast to current complex and over-parametrised architectures that are both compute and time intensive, our method ignores the visual stimuli, ignores the sequencing of dialogue, does not need gradients, uses off-the-shelf feature extractors, has at least an order of magnitude fewer parameters, and learns in practically no time. We argue that these results are indicative of issues in current approaches to visual dialogue and conduct analyses to highlight implicit dataset biases and effects of over-constrained evaluation metrics. Our code is publicly available.
Original languageEnglish
Number of pages6
Publication statusPublished - 7 Dec 2018
EventCritiquing and Correcting Trends in Machine Learning: NeurIPS 2018 Workshop - Montreal, Canada
Duration: 7 Dec 20187 Dec 2018
https://ml-critique-correct.github.io/

Workshop

WorkshopCritiquing and Correcting Trends in Machine Learning
Abbreviated titleCRACT 2018
CountryCanada
CityMontreal
Period7/12/187/12/18
Internet address

Cite this