TY - RPRT TI - Learning Robot Activities from First-Person Human Videos Using Convolutional Future Regression AU - Jangwon Lee AU - Michael S. Ryoo PY - 2017 UR - https://arxiv.org/abs/1703.01040 ID - 1703.01040 ER -