SHD360: A Benchmark Dataset for Salient Human Detection in 360{\\deg}\n Videos

Abstract

Salient human detection (SHD) in dynamic 360{\\deg} immersive videos is of\ngreat importance for various applications such as robotics, inter-human and\nhuman-object interaction in augmented reality. However, 360{\\deg} video SHD has\nbeen seldom discussed in the computer vision community due to a lack of\ndatasets with large-scale omnidirectional videos and rich annotations. To this\nend, we propose SHD360, the first 360{\\deg} video SHD dataset which contains\nvarious real-life daily scenes. Since so far there is no method proposed for\n360{\\deg} image/video SHD, we systematically benchmark 11 representative\nstate-of-the-art salient object detection (SOD) approaches on our SHD360, and\nexplore key issues derived from extensive experimenting results. We hope our\nproposed dataset and benchmark could serve as a good starting point for\nadvancing human-centric researches towards 360{\\deg} panoramic data. The\ndataset is available at https://github.com/PanoAsh/SHD360.\n