Skip to content

This repository contains the research project that enables the robot to automatically join a group based on the modeled personal, social and public spaces of the group using deep reinforcement learning.

Notifications You must be signed in to change notification settings

usr-lab/pepper-social

 
 

Repository files navigation

Pepper Social Scenarios

The pepper social scenarios is implemented using ml-agents. It is still under development.

Unity version: 2018.1.0b13 (beta) Unity ML-Agents version: 0.3.1b

Environments

Pepper robot approaches people: The pepper robot needs to approach people from a right angle:

  • +0.1 if getting closer.
  • -0.05 time penalty.
  • +1.0 when reaching target.

Pepper robot approaches people runtime after training. (current stage)

Project Setup

  • The TensorflowSharp plugins folder was omitted from this project due to the massive file sizes. You will need to import this set of Unity plugins yourself. You can download the TensorFlowSharp plugin as a Unity package here.

Follow Me

Code states

BASELINE-v0 https://github.com/gaoyuankidult/PepperSocial/tree/6a506f3b43f48b881d5d8828e0f781e115a8b155

About

This repository contains the research project that enables the robot to automatically join a group based on the modeled personal, social and public spaces of the group using deep reinforcement learning.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published