Temporal Interpolation Is All You Need for Dynamic Neural Radiance Fields

Sungheon Park, Minjung Son, Seokhwan Jang, Young Chun Ahn, Ji-Yeon Kim, Nahyup Kang

Abstract:

Temporal interpolation often plays a crucial role to learn
meaningful representations in dynamic scenes. In this pa-
per, we propose a novel method to train spatiotemporal neu-
ral radiance fields of dynamic scenes based on temporal
interpolation of feature vectors. Two feature interpolation
methods are suggested depending on underlying represen-
tations, neural networks or grids. In the neural representa-
tion, we extract features from space-time inputs via multi-
ple neural network modules and interpolate them based on
time frames. The proposed multi-level feature interpolation
network effectively captures features of both short-term and
long-term time ranges. In the grid representation, space-
time features are learned via four-dimensional hash grids,
which remarkably reduces training time. The grid repre-
sentation shows more than 100×faster training speed than
the previous neural-net-based methods while maintaining
the rendering quality. Concatenating static and dynamic
features and adding a simple smoothness term further im-
prove the performance of our proposed models. Despite the
simplicity of the model architectures, our method achieved
state-of-the-art performance both in rendering quality for
the neural representation and in training speed for the grid
representation.