Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes
This work addresses the problem of high-quality space-time view synthesis for dynamic scenes, which is important for virtual reality and content creation.
This paper introduces Neural Scene Flow Fields to model dynamic scenes from monocular video, enabling novel view and time synthesis. The method significantly outperforms recent monocular view synthesis approaches.
We present a method to perform novel view and time synthesis of dynamic scenes, requiring only a monocular video with known camera poses as input. To do this, we introduce Neural Scene Flow Fields, a new representation that models the dynamic scene as a time-variant continuous function of appearance, geometry, and 3D scene motion. Our representation is optimized through a neural network to fit the observed input views. We show that our representation can be used for complex dynamic scenes, including thin structures, view-dependent effects, and natural degrees of motion. We conduct a number of experiments that demonstrate our approach significantly outperforms recent monocular view synthesis methods, and show qualitative results of space-time view synthesis on a variety of real-world videos.