The main target of video coding is to maximize the visual quality at the given bit rate budget. With the explosion of video data, efficient video coding technology is becoming more and more urgent. However, in existing coding theory, most image/video quality assessment metrics were related to PSNR or SSIM, which can hardly reflect the human perceptual quality. On the other hand, most existing video coding methods were designed to eliminate the spatial and temporal redundancies without consideration of visual redundancy. It is reasonable to consider QoE of users, human visual characteristics and video content features in the process of designing video coding methods so as to further improve the coding efficiency.

With the development of VR technology, Multi-view/3D video and point cloud have been applied into manufacturing, sports, design, etc., since they are capable of representing high quality 3D world scene, such as 3D depth impression and interactive selection of arbitrary viewpoint/direction within a certain range of distances. Considering these features and with the technological advancements in display technology, it would allow many new visual media applications, such as photorealistic rendering of 3D scenes, free-viewpoint television (FTV), 3D television (3DTV) broadcasting, and also include mass customization point cloud applications, to provide exciting visual experience for users. Due to the various applications, point cloud compression of static data aims to exploit the spatial redundancy, point cloud streaming aims to achieve a balance between the visual quality and bulky data on display devices.

Immersive Virtual Reality (VR) can provide high definition, more vivid and interactive visual experience, which has a large number of potential market demands. However, it still faces many challenges, including: 1) video content generation, high quality video capturing and video stitching technology; 2) low complexity while high quality video rendering technology and deep user interaction; 3) visual perception models to improve user's QoE.

Intelligent Video Analytics (IVA) products add artificial intelligence to cameras for analyzing video content in real-time, extracting metadata, sending out alerts and providing actionable intelligence to security surveillance or other systems. Video analytics can be embedded at the edge (in-camera), in server’s on-premise, and/or on-cloud. Machine learning and, deep learning approaches, have revolutionized video analytics, which include behavior recognition, image classification, detection and tracking etc. Based on machine learning and deep learning methods, IVA software can contribute in a major way by providing a means of accurately dealing with volumes of information.