SD Study Group

From AHC Wiki
Revision as of 12:15, 5 February 2018 by Tomomasa-f (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

This is a page for the NAIST Spoken Dialogue study group, a study group at the NAIST Information Science Department. It is held Friday at 15:10. If you are interested in joining, please contact Yoshino for more details.

NAIST音声対話勉強会のサイトです。 勉強会は金曜日の15:10に開かれます。興味があれば、吉野までご連絡ください。

Upcoming Meetings / 今期の勉強会

  • This is a draft schedule. If there are any problem with it, please mail to [sugiyama.kyoshiro.sc7(at)].
Date Presenter Contents Slides
1/12 Shinagawa Learning Cooperative Visual Dialog Agents with Deep Reinforcement Learning (ICCV2017) PDF
1/26 Wakimoto Unstructured Human Activity Detection from RGBD Images PDF
2/2 Katsumi Grounded language learning in a simulated 3D world ( PDF
2/9 Furukawa Attentive listening system with backchanneling, response generation and flexible turn-taking (SIGDIAL 2017) PDF
2/16 Sugiyama
2/23 Takahashi

Previous Meetings / 今までの勉強会

Date Presenter Contents Slides
6/30 Shinagawa 1. Introduction PDF
7/7 Katsumi,Nakayama 2. Learning Basics and Linear Models PDF
3. From Linear Models to Multi-layer Perceptions PDF
7/14 Hayashi 4. Feed-forward Neural Networks PDF
Furukawa 5. Neural Network Training PDF
7/21 Wakimoto 6. Features for Textual Data PDF
Ikeuchi 7. Case Studies of NLP Features PDF
7/28 Takahashi 8. From Textual Features to Inputs PDF
Yen 9. Language Modeling PDF
8/5 Kawano,Murase 10. Pre-trained Word Representations MURASEKAWANO
TBD Toyoshima 11. Using Word Embeddings PDF
TBD Shinagawa 12. Case Study: A Feed-forward Architecture for Sentence Meaning Inference PDF
TBD Tung 13. Ngram Detectors: Convolutional Neural Networks PDF
TBD Murase 14. Recurrent Neural Networks: Modeling Sequences and Stacks PDF
TBD Kawano 15. Concrete Recurrent Neural Network Architectures PDF
TBD Toyoshima 16. Modeling with Recurrent Networks PDF
TBD Ikuta 17. Conditioned Generation PDF
TBD Sugiyama 18. Modeling Trees with Recursive Neural Networks
19. Structured Output Prediction
TBD Shinagawa 20. Cascaded, Multi-task and Semi-supervised Learning

2017 Book reading

  • A:対話システム(中野幹生, 駒谷和範, 船越孝太郎, 中野有紀子. コロナ社, 2015)
Date Presenter Contents Slides
4/28 Yoshino 音声対話システムの概要 PDF
5/12 Furukawa A-2.1~2.3 対話のモデルー基本構造と共通基盤 PDF
5/19 Ikeuchi A-2.4~2.5 プランと背景構造 PDF
Hayashi A-3.1~3.3 対話管理
5/26 Ujiro A-3.4~3.5 対話の主導権と入出力
Wakimoto A-5.1 統計的意図理解
6/8 Wakimoto A-5.1 統計的意図理解 (continued)
Katsumi A-5.2 適応的な対話管理・応答生成 PDF
Canceled A-5.3~5.4 問題検出と話者交代
Canceled A-5.5~5.6 マルチモーダル・HCI

2016 Paper reading

Date Presenter Contents Slides
10/20 Yoshino Towards End-to-End Learning for Dialog State Tracking and Management using Deep Reinforcement Learning (In Proc SIGDIAL 2016) Slides
10/27 Mizukami A Wizard-of-Oz Study on A Non-Task-Oriented Dialog Systems That Reacts to User Engagement (in Proc. SIGDIAL2016) Slides
11/10 Shinagawa Introduction of LSTM PDF_ja,pptx_ja


11/24 Sugiyama Strategy and Policy Learning for Non-Task-Oriented Conversational System PDF
12/1 Ishikawa Are you convinced? A Wizard of Oz study to test emotional vs. rational persuasion strategies in dialogues PDF
12/8 Sasano Progressive Neural Networks PDF
12/15 O-uchi (Matsumoto lab.)
12/22 Morimoto (Matsumoto lab.)
1/5 Nunu

2016 Book Reading

Date Presenter Contents Slides
5/12 Yoshino B-1 音声対話システムの概要 EN JP
5/19 Murase A-2.1~2.3 対話のモデルー基本構造と共通基盤 JP
Osamura.A A-2.4~2.5 プランと背景構造 JP
5/26 Kawano A-3.1~3.3 対話管理 JP
Hosomi A-3.4~3.5 対話の主導権と入出力 JP
6/2 Toyoshima A-5.1 統計的意図理解 JP
Volunteer A-5.2 適応的な対話管理・応答生成 JP
6/9 Yanagita A-5.3~5.4 問題検出と話者交代


Mori A-5.5~5.6 マルチモーダル・HCI JP
6/16 Ishikawa D-1 Background of emotional system EN
Nurul D-2 Review of emotion theories EN
6/23 Mizukami E-1 Background of task oriented dialogue system EN
Tung E-2.1~2.2.3 Reinforcement learning based dialogue management EN
6/30 Shinagawa E-2.2.4~2.2.6 POMDP based dialogue management EN.pptx EN.pdf
Sasano E-2.3 User simulator EN.pdf
7/14 Ido C-1 Background of non-goal oriented spoken dialogue system EN
Sugiyama C-2 Statistical learning of domain-dependent semantic structure EN
 ? Ouchi, Morimoto? C-5 Statistical learning of dialogue model by tracking dialogue state and user focus

Paper Reading

Date Loc. Content Jornal/Conference Presenter Slides
2015/12/9 AHC-Lab Best of both worlds: human-machine collaboration for object annotation CVPR2015 Shinagawa slides,Japanese
2016/2/9 AHC-Lab An Analysis-By-Synthesis Approach to Multisensory Object Shape Perceptio (not open access) NIPS2015 Workshop Sasano slides
2016/2/16 AHC-Lab From Sensory Signals to Modality-Independent Conceptual Representations: A Probabilistic Language of Thought Approach PLoS Computational Biology Sasano slides
2016/2/23 AHC-Lab Don't Grin When You Win: The Social Costs of Positive Emotion Expression in Performance Situations (not open access) Emotion Ishikawa slides
2016/3/1 AHC-Lab Adapting to Multiple Affective States in Spoken Dialogue SIGDIAL Nurul Slides
2016/3/8 AHC-Lab Exploiting knowledge base to generate responses for natural language dialog listening agents SIGDIAL Sugiyama Slides
2016/3/15 AHC-Lab [] SEMDIAL Tung Slides

Book Reading


Englich:Spoken Dialogue Systems, Kristina Jokinen and Michael McTear. Morgan & Claypool.

Date Loc. Content Presenter Slides
2015/5/25 AHC-Lab 対話のモデル, dialogue modeling Ishikawa, Shinagawa Ishikawa,Shinagawa
2015/6/8 AHC-Lab 対話システムの構成と処理の概要, architecture of dialogue system (English chapter: 1.1, 1.2, 2.1, 2.2, 2.3) Mizukami, Sugiyama Mizukami,Sugiyama
2015/6/15 AHC-Lab 対話システムの設計と構築, How to construct a dialogue system (English chapter: 1.4, 1.5, 2.4, 3.1, 3.2, 4, 6) Lasguido Lasguido
2015/6/29 AHC-Lab 対話システムの設計と構築(続き), How to construct a dialogue system (continue from previous study) Nurul, Maeda Nurul,Maeda
2015/7/13 AHC-Lab 対話システムの発展技術, Utilities of dialogue system (English chapter: 1.3, 1.6, 5) Sasano,Hiraoka Sasano