Presentation 2017-08-21
Meeting Extracts for Group Discussions using Multimodal Convolutional Neural Networks
Fumio Nihei, Yukiko Nakano, Yutaka Takase,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) With the goal of extracting meeting minutes from group discussion corpus, this study proposes multimodal fusion models by employing Convolutional Neural Networks (CNNs). First, unimodal models are created using raw behavioral data such as speech, head motion, and face tracking. These models are then integrated into a fusion model that works as a classifier. The main advantage of this work is that the proposed models were trained without any hand-crafted features, and in most cases they outperformed a baseline model trained using hand-crafted features. It was also found that multimodal fusion is useful in applying the CNN approach to model multimodal multiparty interaction.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Meeting extracts / Convolutional neural network / Multimodal fusion
Paper # HCS2017-57
Date of Issue 2017-08-13 (HCS)

Conference Information
Committee HCS
Conference Date 2017/8/20(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Seikei University
Topics (in Japanese) (See Japanese page)
Topics (in English) Support technologies based on communication sensing, etc.
Chair Yukiko Nakano(Seikei Univ.)
Vice Chair Nobuyuki Watanabe(Kanazawa Inst. of Tech.) / Tomoo Inoue(Univ. of Tsukuba)
Secretary Nobuyuki Watanabe(Ritsumeikan Univ.) / Tomoo Inoue(Osaka Electro-Comm. Univ.)
Assistant Kazuki Takashima(Tohoku Univ.) / Ken Fujiwara(Osaka Univ. of Economic) / Kazunori Terada(Gifu Univ.) / Atsushi Kimura(Nihon Univ.)

Paper Information
Registration To Technical Committee on Human Communication Science
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Meeting Extracts for Group Discussions using Multimodal Convolutional Neural Networks
Sub Title (in English)
Keyword(1) Meeting extracts
Keyword(2) Convolutional neural network
Keyword(3) Multimodal fusion
1st Author's Name Fumio Nihei
1st Author's Affiliation Seikei University(Seikei Univ.)
2nd Author's Name Yukiko Nakano
2nd Author's Affiliation Seikei University(Seikei Univ.)
3rd Author's Name Yutaka Takase
3rd Author's Affiliation Seikei University(Seikei Univ.)
Date 2017-08-21
Paper # HCS2017-57
Volume (vol) vol.117
Number (no) HCS-177
Page pp.pp.55-59(HCS),
#Pages 5
Date of Issue 2017-08-13 (HCS)