Neural Radiance Field (NeRF) approaches learn the un-
derlying 3D representation of a scene and generate photo-
realistic novel views with high fidelity. However, most pro-
posed settings concentrate on modelling a single object or
a single level of a scene. However, in the real world, we
may capture a scene at multiple levels, resulting in a lay-
ered capture. For example, tourists usually capture a mon-
ument’s exterior structure before capturing the inner struc-
ture. Modelling such scenes in 3D with seamless switch-
ing between levels can drastically improve immersive ex-
periences. However, most existing techniques struggle in
modelling such scenes. We propose Strata-NeRF, a single
neural radiance field that implicitly captures a scene with
multiple levels. Strata-NeRF achieves this by condition-
ing the NeRFs on Vector Quantized (VQ) latent represen-
tations which allow sudden changes in scene structure. We
evaluate the effectiveness of our approach in multi-layered
synthetic dataset comprising diverse scenes and then fur-
ther validate its generalization on the real-world RealEstate
10k dataset. We find that Strata-NeRF effectively captures
stratified scenes, minimizes artifacts, and synthesizes high-
fidelity views compared to existing approaches.
