Strata-NeRF : Neural Radiance Fields for Stratified Scenes
This work addresses the challenge of immersive 3D scene modeling for applications like tourism or virtual experiences, though it appears incremental as it builds on existing NeRF methods with a novel conditioning approach.
The paper tackles the problem of modeling 3D scenes with multiple levels (e.g., exterior and interior structures) using Neural Radiance Fields (NeRF), which existing methods struggle with, and proposes Strata-NeRF, a single NeRF conditioned on Vector Quantized latent representations to capture such stratified scenes, achieving high-fidelity view synthesis with minimized artifacts compared to existing approaches.
Neural Radiance Field (NeRF) approaches learn the underlying 3D representation of a scene and generate photo-realistic novel views with high fidelity. However, most proposed 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 layered capture. For example, tourists usually capture a monument's exterior structure before capturing the inner structure. Modelling such scenes in 3D with seamless switching between levels can drastically improve immersive experiences. 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 conditioning the NeRFs on Vector Quantized (VQ) latent representations which allow sudden changes in scene structure. We evaluate the effectiveness of our approach in multi-layered synthetic dataset comprising diverse scenes and then further validate its generalization on the real-world RealEstate10K dataset. We find that Strata-NeRF effectively captures stratified scenes, minimizes artifacts, and synthesizes high-fidelity views compared to existing approaches.