2021-07-01-bar21a.md

title	abstract	layout	series	publisher	issn	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	container-title	volume	genre	issued	pdf	extras
Compositional Video Synthesis with Action Graphs	Videos of actions are complex signals containing rich compositional structure in space and time. Current video generation methods lack the ability to condition the generation on multiple coordinated and potentially simultaneous timed actions. To address this challenge, we propose to represent the actions in a graph structure called Action Graph and present the new "Action Graph To Video" synthesis task. Our generative model for this task (AG2Vid) disentangles motion and appearance features, and by incorporating a scheduling mechanism for actions facilitates a timely and coordinated video generation. We train and evaluate AG2Vid on CATER and Something-Something V2 datasets, which results in videos that have better visual quality and semantic consistency compared to baselines. Finally, our model demonstrates zero-shot abilities by synthesizing novel compositions of the learned actions.	inproceedings	Proceedings of Machine Learning Research	PMLR	2640-3498	bar21a	0	Compositional Video Synthesis with Action Graphs	662	673	662-673	662	false	Bar, Amir and Herzig, Roei and Wang, Xiaolong and Rohrbach, Anna and Chechik, Gal and Darrell, Trevor and Globerson, Amir	given family Amir Bar given family Roei Herzig given family Xiaolong Wang given family Anna Rohrbach given family Gal Chechik given family Trevor Darrell given family Amir Globerson	2021-07-01		Proceedings of the 38th International Conference on Machine Learning	139	inproceedings	date-parts 2021 7 1	http://proceedings.mlr.press/v139/bar21a/bar21a.pdf