Natural Language Processing / Speech Processing
Understand human language.
Natural language is the most intuitive interface for most people and situations. By advancing human-robot interaction (HRI) via language, we aim to achieve “robots for everyone”, and a world where even non-experts can make use of such transformative technology. Our research in HRI spans multiple modalities and combines research in NLP, scene understanding, speech recognition, robotics, and augmented reality, in order to provide robust understanding of a wide variety of interactions.
VocabEncounter: NMT-powered Vocabulary Learning by Presenting Computer-Generated Usages of Foreign Words into Users’ Daily Lives
By : Riku Arakawa, Hiromu Yakura, Sosuke Kobayashi
Instance-Based Neural Dependency Parsing
By : Hiroki Ouchi, Jun Suzuki, Sosuke Kobayashi, Sho Yokoi, Tatsuki Kuribayashi, Masashi Yoshikawa, Kentaro Inui
SHAPE: Shifted Absolute Position Embedding for Transformers
By : Shun Kiyono, Sosuke Kobayashi, Jun Suzuki, Kentaro Inui
All Word Embeddings from One Embedding
By : Sho Takase, Sosuke Kobayashi