Under the abundance of applications in computer vision and graphics, neural rendering is emerged as avant-garde technology for generating realistic image and video data and accomplishing a controllable, realizable model of a 3D scene.

Neural rendering is a new research area that integrates rapidly moving technological advances with the unavailability of social science research on their implications for individuals, groups, institutes, and society. Neural rendering conducts deep image and video generation that authorize control of scene properties.

Recently, deep generative models have been widely used for the automatic synthesis of photo-realistic images and videos with neural rendering. Impressive neural rendering applications are Semantic Photo Synthesis and Manipulation, Novel View Synthesis for Objects and Scenes, Free Viewpoint Videos, Learning to Relight, Facial Reenactment, and Body Reenactment.

Open challenges which lead to opportunities for further advancing neural rendering are Generalization, Editability, Scalability, Multimodal Neural Scene Representations, Seamless Integration and Usage, and Multi-Modal Learning. Various neural rendering-related surveys have been introduced and describe useful details such as the recent trends, applications of neural rendering, social implications, challenges, and open research problems.