TY - RPRT TI - Generalizable task representation learning from human demonstration videos: a geometric approach AU - Jun Jin AU - Martin Jagersand PY - 2022 UR - https://arxiv.org/abs/2202.13604 ID - 2202.13604 ER -