学生との共著論文が #Interspeech2026 に採択されました.Spoken DSTの研究です.Interspeechは初参加!
Tomoya Higuchi and Michimasa Inaba: Incremental End-to-End Spoken Dialogue State Tracking with a Multimodal LLM and Reinforcement Learning
We’re excited to announce the release of MultilingualKokoroChat, a multilingual counseling dialogue dataset, along with our paper, accepted at #LREC2026!
https://t.co/MbdYCibCya
https://t.co/ykT9ZLUDnC
Our paper “Emotion Transcription in Conversation: A Benchmark for Capturing Subtle and Complex Emotional States through Natural Language”, accepted at #LREC2026, is now available on arXiv! https://t.co/kWzZGvjEls
論文紹介 3
2名がテキスト対話を行いながらストーリーの共同創作を行うことで収集した対話データセットを構築し,作業者の性格特性・発話数・編集ログに焦点を当てた分析を行いました.
Collecting and Analyzing Human–Human Co-Creation Dialogues for Personalized Creative Support
論文紹介 2
話者の心情を自然言語で記述するタスク「対話における心情記述」の提案とそのためのデータセットを構築しました.
Emotion Transcription in Conversation: A Benchmark for Capturing Subtle and Complex Emotional States through Natural Language
Paper accepted to LREC 2026 !!
Emotion Transcription in Conversation: A Benchmark for Capturing Subtle and Complex Emotional States through Natural Language (Tanaka)
This is a joint work with Inaba-lab in UEC🙇♀️