Efficient Attention Mechanism for Visual Dialog that Can Handle All the Interactions Between Multiple Inputs

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

It has been a primary concern in recent studies of vision and language tasks to design an effective attention mechanism dealing with interactions between the two modalities. The Transformer has recently been extended and applied to several bi-modal tasks, yielding promising results. For visual dialog, it becomes necessary to consider interactions between three or more inputs, i.e., an image, a question, and a dialog history, or even its individual dialog components. In this paper, we present a neural architecture named Light-weight Transformer for Many Inputs (LTMI) that can efficiently deal with all the interactions between multiple such inputs in visual dialog. It has a block structure similar to the Transformer and employs the same design of attention computation, whereas it has only a small number of parameters, yet has sufficient representational power for the purpose. Assuming a standard setting of visual dialog, a layer built upon the proposed attention block has less than one-tenth of parameters as compared with its counterpart, a natural Transformer extension. The experimental results on the VisDial datasets validate the effectiveness of the proposed approach, showing improvements of the best NDCG score on the VisDial v1.0 dataset from 57.59 to 60.92 with a single model, from 64.47 to 66.53 with ensemble models, and even to 74.88 with additional finetuning.

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2020 - 16th European Conference, 2020, Proceedings
EditorsAndrea Vedaldi, Horst Bischof, Thomas Brox, Jan-Michael Frahm
PublisherSpringer Science and Business Media Deutschland GmbH
Pages223-240
Number of pages18
ISBN (Print)9783030585853
DOIs
Publication statusPublished - 2020
Event16th European Conference on Computer Vision, ECCV 2020 - Glasgow, United Kingdom
Duration: 2020 Aug 232020 Aug 28

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12369 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference16th European Conference on Computer Vision, ECCV 2020
CountryUnited Kingdom
CityGlasgow
Period20/8/2320/8/28

Keywords

  • Attention
  • Multimodality
  • Visual dialog

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Efficient Attention Mechanism for Visual Dialog that Can Handle All the Interactions Between Multiple Inputs'. Together they form a unique fingerprint.

Cite this