Computer Vision and NLU

	Video Fill In the Blank using LR/RL LSTMs with Spatial-Temporal Attentions Amir Mazaheri, Dong Zhang and Mubarak Shah