Egok360: A 360 Egocentric Kinetic Human Activity Video Dataset

Abstract

Recently, there has been a growing interest in wearable sensors which provides new research perspectives for 360 video analysis. However, the lack of 360 datasets in literature hinders the research in this field. To bridge this gap, in this paper we propose a novel Egocentric (first-person) 360 Kinetic human activity video dataset (EgoK360). The EgoK360 dataset contains annotations of human activity with different sub-actions, e.g., activity Ping-Pong with four sub-actions which are pickup-ball, hit, bounce-ball and serve. To the best of our knowledge, EgoK360 is the first dataset in the domain of first-person activity recognition with a 360 environmental setup, which will facilitate the egocentric 360 video understanding. We provide experimental results and comprehensive analysis of variants of the two-stream network for 360 egocentric activity recognition. The EgoK360 dataset can be downloaded from https://egok360.github.io/.

Tasks

Reproductions