[Seminar] "End to end approach for behavior generation in robot systems" by Prof. Tetsuya Ogata
Prof. Tetsuya Ogata, Waseda University
End to end approach for behavior generation in robot systems
In this talk, I will present the end to end learning approach on deep neural network models, which enable robot systems to recognize the environment and to interact with human beings. Our model introduces the combination of a convolution neural model and a recurrent neural model which enable a humanoid robot NEXTAGE to manipulate the various objects including soft materials. We also investigate a robot manipulation model that uses DNNs and is able to execute long sequential dynamic tasks by performing multiple short sequential tasks at appropriate times. The internal state of the MTRNN is constrained to have similar values at the initial and final steps of a motion, allowing the motions to be differentiated based on the initial image input. The future problems for the robot application will be discussed in the end of this talk.