Contributed talk AVID: Translating Human Demonstrations for Automated Learning