The YTTM is a cognitive model of multimodal realtime turntaking that has been tested in numerous two-party dialogue systems. This video shows an extension to the model that allows it to be used to control agents in realtime multimodal dialogue. Up to 12 agents have been tested.
Thanks for the positive feedback -- I appreciate it!
The work is based on over a decade's worth of interaction and dialogue research, so yes, I hope it represents some progress :-)
cheers,
thorisson
thorisson 1 year ago
This is some very interesting Research. I think this is definitely progress so I will be promoting to the community at Thoughtware (dot) TV.
Thanks for posting!
thoughtwaretv 1 year ago