Learning to Segment Actions from Observation and Narration

Daniel Fried, Jean-Baptiste Alayrac, Phil Blunsom, Chris Dyer, Stephen Clark, Aida Nematzadeh


Abstract
We apply a generative segmental model of task structure, guided by narration, to action segmentation in video. We focus on unsupervised and weakly-supervised settings where no action labels are known during training. Despite its simplicity, our model performs competitively with previous work on a dataset of naturalistic instructional videos. Our model allows us to vary the sources of supervision used in training, and we find that both task structure and narrative language provide large benefits in segmentation quality.
Anthology ID:
2020.acl-main.231
Volume:
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2020
Address:
Online
Editors:
Dan Jurafsky, Joyce Chai, Natalie Schluter, Joel Tetreault
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2569–2588
Language:
URL:
https://aclanthology.org/2020.acl-main.231
DOI:
10.18653/v1/2020.acl-main.231
Bibkey:
Cite (ACL):
Daniel Fried, Jean-Baptiste Alayrac, Phil Blunsom, Chris Dyer, Stephen Clark, and Aida Nematzadeh. 2020. Learning to Segment Actions from Observation and Narration. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 2569–2588, Online. Association for Computational Linguistics.
Cite (Informal):
Learning to Segment Actions from Observation and Narration (Fried et al., ACL 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/landing_page/2020.acl-main.231.pdf
Video:
 http://slideslive.com/38929315
Code
 dpfried/action-segmentation
Data
CrossTask