Actor and Observer: Joint Modeling of First and Third-Person Videos

TL;DR

This paper introduces Charades-Ego, a large dataset of paired first- and third-person videos, enabling joint modeling of perspectives and improving human action recognition through weakly supervised learning.

Contribution

The paper presents the Charades-Ego dataset and demonstrates a novel approach to jointly model first- and third-person videos for better action recognition.

Findings

Effective transfer of knowledge from third- to first-person videos

Joint representation improves action recognition accuracy

Dataset enables new research in egocentric vision

Abstract

Several theories in cognitive neuroscience suggest that when people interact with the world, or simulate interactions, they do so from a first-person egocentric perspective, and seamlessly transfer knowledge between third-person (observer) and first-person (actor). Despite this, learning such models for human action recognition has not been achievable due to the lack of data. This paper takes a step in this direction, with the introduction of Charades-Ego, a large-scale dataset of paired first-person and third-person videos, involving 112 people, with 4000 paired videos. This enables learning the link between the two, actor and observer perspectives. Thereby, we address one of the biggest bottlenecks facing egocentric vision research, providing a link from first-person to the abundant third-person data on the web. We use this data to learn a joint representation of first and third-person videos, with only weak supervision, and show its effectiveness for transferring knowledge from the third-person to the first-person domain.