Introduction to the Issue on Perception-Driven 360° Video Processing

Recent years have witnessed the rapid development of virtual reality (VR). Above 90% of VR content is in the form of 360° video, also called omnidirectional video or panoramic video. Generally speaking, 360° video offers immersive and interactive viewing experience, as the viewers are able to freely move their heads in the range of 360° × 180° to access different viewports. Therefore, the perception of 360° video is significantly different from that of traditional video. It is thus necessary to study the perception model of 360° video. On the other hand, the perception model can be used to significantly remove the redundancy of 360° video by predicting viewports and eye fixations of the viewers. Due to this, the perception model plays a crucial role in improving efficiency and efficacy of 360° video processing, such as compression and visual quality assessment (VQA).

To bridge the gap between the perception model and application for 360° video, this special issue released Call for Papers outlining the topics on the state-of-the-art research, innovations, and applications of perception-driven 360° video processing. Then, this special issue attracted many submissions in the filed of 360° video processing, in particular perception-driven signal processing methods for 360° videos or images. Afterwards, the Guest Editors handled a rigorous review process, and 16 papers were finally accepted and published in this special issue.