Suraj Pattar

Robotics AI Research Engineer / Ph.D. in AI and Robotics / Data Scientist / 3D Printing Enthusiast

Adaptive Multi-Party Human Robot Interaction

Keywords: Human Robot Interaction, Human Activity Recognition, Machine Learning, Deep Learning,Unsupervised Learning, Time Series Data Classification, Face Recognition, Chatbot Development

Project Description

Github Project

This was my Master Thesis Project at GV Lab, Tokyo University of Agriculture and Technology, Tokyo.

The goal of the research was to build a Personlized Human Robot Interaction system. We used time-series data from Kinect to build an Auto-encoder for Human Activity Recognition. We also made use of Facial Recognition and Chatbot for achieving our goals.

Platforms:

Ubuntu (Linux), Windows 10

Software:

Python, C#, Bash

Technology:

Tensorflow, Scikit-learn, Dialogflow, Choreographe.

Hardware:

Pepper, Kinect, External microphone, Tablet.

Personalized Interaction with Pepper Robot

Dialogflow chatbot integrated with Pepper Robot

Using Facial Recognition for Personalized Interaction

Pepper greets using kNN trained Face Recognition

Pepper greets using One Shot Face Recognition