Research
Example of dialogue (Japanese)
Left : Bot / Right : Human
Dialogue System
We are working on the development of a active listening system using Twitter data.
Active Listening is a conversational technique in which you do not just listen to the other person, but actively listen while asking questions as appropriate.
This system can be a conversation partner for the elderly or people who live alone and have no one to talk to.
Multimodal Dialogue System
We are working on the development of multimodal dialogue systems that use not only verbal information such as text, but also non-verbal information such as posture and gestures.
This is a prototype of a multimodal dialogue system using "Unity-chan" (© UTJ/UCL).
The image is projected on the naked-eye stereoscopic display ELF-SR1. By recognizing the speaker's face position, the system speaks to the speaker.
Featured in the University News
The 3D model in the background is a reproduction of the Cybernetics Laboratory, to which I belong, and was created by volunteers from the laboratory.
The virtual lab using this 3D model is available on cluster.
VR Multimodal Dialogue System
This video is a VR reproduction of the dialogue with the chatbot above.
You can experience the dialogue in a 360-degree video.
If you have VR equipment, you can even enjoy it in 3D sound!
Example of question generation (Japanese)
Question Generation
We are working on developing a Case Frame Compilation-based question generation system.
This research will enable rapid and high-volume generation of multiple questions from a single input sentence.