Our research interest is on a machine intelligence which interacts with the dynamic environment. We are challenging this theme by taking constructivist approaches with robot systems and neural network model (deep learning). The approach includes imitation learning, linguistic-behavioral integration, human-robot cooperation, and multi-modal active sensing.
Translation from short sentence with logic words to robot motion (2016) Exhibited at NIPS2016
Folding towels at Artificial Intelligence Research Center, AIST (2016) Exhibited at Cebit2017, CEATEC2017, iREX2017
Multi-modal AI robot with Exawizards, Denso-wave, and Beckhoff Automations (2017) Exhibited at iREX2017, NVIDIA GTC Japan 2017, Science Robotics Meeting 2018
Whole body motion of door approaching, opening and passing with Hitachi co. (2018) (You need to login Facebook to see the movie.) Exhibited at NVIDIA GTC Japan 2018
Weighing by Exawizards and Denso co. (2018) Exhibited at NVIDIA GTC Japan 2018, World Robot Exhibition 2018, ROBODEX2019
Palletizing by Exawizards and Denso co. (2018) Exhibited at NVIDIA GTC Japan 2018, World Robot Exhibition 2018, ROBODEX2019
The Robotic Technology Middleware (RT-middleware) is a middleware that will boost the usability of RT technology elements, like Actuators, Sensors, and Controller Softwares. We are applying those technologies for the activities in Interactive Media Arts.