A Portable Multiscopic Camera for Novel View and Time Synthesis in Dynamic Scenes
This addresses the problem of rendering dynamic scenes from arbitrary viewpoints and times for applications like virtual reality or film production, representing a novel method for a known bottleneck.
The researchers tackled novel view and time synthesis in dynamic scenes using a portable multiscopic camera system, developing a model that maps 6D coordinates to render high-quality images from any viewpoint and time, with experiments showing it outperforms alternative solutions qualitatively and quantitatively.
We present a portable multiscopic camera system with a dedicated model for novel view and time synthesis in dynamic scenes. Our goal is to render high-quality images for a dynamic scene from any viewpoint at any time using our portable multiscopic camera. To achieve such novel view and time synthesis, we develop a physical multiscopic camera equipped with five cameras to train a neural radiance field (NeRF) in both time and spatial domains for dynamic scenes. Our model maps a 6D coordinate (3D spatial position, 1D temporal coordinate, and 2D viewing direction) to view-dependent and time-varying emitted radiance and volume density. Volume rendering is applied to render a photo-realistic image at a specified camera pose and time. To improve the robustness of our physical camera, we propose a camera parameter optimization module and a temporal frame interpolation module to promote information propagation across time. We conduct experiments on both real-world and synthetic datasets to evaluate our system, and the results show that our approach outperforms alternative solutions qualitatively and quantitatively. Our code and dataset are available at https://yuenfuilau.github.io.