Cross-Task Weakly Supervised Learning From Instructional Videos

Dimitri Zhukov, Jean-Baptiste Alayrac, Ramazan Gokberk Cinbis, David Fouhey, Ivan Laptev, Josef Sivic; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 3537-3545


In this paper we investigate learning visual models for the steps of ordinary tasks using weak supervision via instructional narrations and an ordered list of steps instead of strong supervision via temporal annotations. At the heart of our approach is the observation that weakly supervised learning may be easier if a model shares components while learning different steps: "pour egg" should be trained jointly with other tasks involving "pour" and "egg". We formalize this in a component model for recognizing steps and a weakly supervised learning framework that can learn this model under temporal constraints from narration and the list of steps. Past data does not permit systematic studying of sharing and so we also gather a new dataset aimed at assessing cross-task sharing. Our experiments demonstrate that sharing across tasks improves performance, especially when done at the component level and that our component model can parse previously unseen tasks by virtue of its compositionality.

Related Material

author = {Zhukov, Dimitri and Alayrac, Jean-Baptiste and Cinbis, Ramazan Gokberk and Fouhey, David and Laptev, Ivan and Sivic, Josef},
title = {Cross-Task Weakly Supervised Learning From Instructional Videos},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2019}