Social Relation Recognition in Egocentric Photostreams
This addresses social relation recognition for users of wearable cameras, but appears incremental as it builds on existing social theory and deep learning approaches.
The paper tackles the problem of automatically categorizing social interactions from egocentric camera footage by proposing a deep learning architecture that uses hierarchical label structure and frame-level social attributes. Experimental results on the new EgoSocialRelation dataset demonstrate the method's effectiveness, though no concrete performance numbers are provided.
This paper proposes an approach to automatically categorize the social interactions of a user wearing a photo-camera 2fpm, by relying solely on what the camera is seeing. The problem is challenging due to the overwhelming complexity of social life and the extreme intra-class variability of social interactions captured under unconstrained conditions. We adopt the formalization proposed in Bugental's social theory, that groups human relations into five social domains with related categories. Our method is a new deep learning architecture that exploits the hierarchical structure of the label space and relies on a set of social attributes estimated at frame level to provide a semantic representation of social interactions. Experimental results on the new EgoSocialRelation dataset demonstrate the effectiveness of our proposal.