Impact of User's Movements in an Immersive 360° Video Streaming with Differentiated Compression

Nowadays, Virtual Reality (VR) is becoming a crucial technology in various sectors, emerging as one of the fastest-growing market segments in recent years. Moreover, continuous advancements in networking and multimedia technologies are increasing the interest regarding several VR applications. Given that 360° video transmission is bandwidth hungry, video compression plays a key role for the deployment of applications based on it. Since users have a partial view of the whole 360° image stream, that is the portion they are focusing, while the part that is behind them is not seen, a widely adopted way to compress this kind of stream is by using a differentiated compression, with higher quality for the image areas that mostly influence the perceived quality. However, at the best of our knowledge, perceived quality has been never correlated to the grade of interaction between the user and the viewed scene, and to the network delay. To this purpose, the main objective of this paper is proposing a 360° video streaming framework with differentiated encoding, and analyze the impact of the user's movements while interacting with the viewed scene, in different cases of network delay.