On the Role of Event Boundaries in Egocentric Activity Recognition from Photostreams
This work addresses activity recognition for egocentric photostreams, an emerging domain, but is incremental as it builds on existing boundary detection methods.
The paper investigates the impact of automatically computed event boundaries on activity recognition in egocentric photostreams, showing that this approach can improve results, and introduces a new annotated dataset from 15 users to demonstrate generalization of deep learning models to unseen users.
Event boundaries play a crucial role as a pre-processing step for detection, localization, and recognition tasks of human activities in videos. Typically, although their intrinsic subjectiveness, temporal bounds are provided manually as input for training action recognition algorithms. However, their role for activity recognition in the domain of egocentric photostreams has been so far neglected. In this paper, we provide insights of how automatically computed boundaries can impact activity recognition results in the emerging domain of egocentric photostreams. Furthermore, we collected a new annotated dataset acquired by 15 people by a wearable photo-camera and we used it to show the generalization capabilities of several deep learning based architectures to unseen users.