About me:

• I'm researching natural language processing and computational linguistics, particularly Japanese language, multilingual and code-switching ASR and multimodal models and machine translation at Tokyo-based Reazon Human Interaction Lab
  • We most recently released a new version of ReazonSpeech, one of the most accurate Japanese ASR models to date
  • We try to open source most, if not all, of our work
  • We're also working on a fully open-source humanoid robot arm, called OpenArm
• I'm a part of the Computational Social Listening Lab at Penn, where I work in computational social science and NLP
  • My current project is about using NLP methods to uncover insights about social norms across cultures
• I am interested in machine learning, specifically computer vision and natural language processing, and previously interned at NVIDIA under the DeepStream team, where I worked on multi-object tracking algorithms and low level optimization. I also conducted machine translation research at the Miyao research group of the University of Tokyo, where I researched the effect of Japanese case markers on the robustness of Japanese-English MT. I find low level and systems level work interesting too
• 📫 You can reach me via LinkedIn here or [email protected]