Incremental Multi-Scene Modeling via Continual Neural Graphics Primitives

TL;DR

This paper introduces C-NGP, a continual learning framework that incrementally encodes multiple 3D scenes into a single NeRF model, maintaining high rendering quality without increasing parameters.

Contribution

It presents a novel continual learning method for NeRFs that integrates multiple scenes into one model using generative replay, avoiding parameter growth.

Findings

C-NGP can model 8 scenes simultaneously with minimal quality loss.

C-NGP maintains high-quality novel-view rendering on synthetic and real datasets.

The approach enables multiple style edits within a single network.

Abstract

Neural radiance fields (NeRF) have revolutionized photorealistic rendering of novel views for 3D scenes. Despite their growing popularity and efficiency as 3D resources, NeRFs face scalability challenges due to the need for separate models per scene and the cumulative increase in training time for multiple scenes. The potential for incrementally encoding multiple 3D scenes into a single NeRF model remains largely unexplored. To address this, we introduce Continual-Neural Graphics Primitives (C-NGP), a novel continual learning framework that integrates multiple scenes incrementally into a single neural radiance field. Using a generative replay approach, C-NGP adapts to new scenes without requiring access to old data. We demonstrate that C-NGP can accommodate multiple scenes without increasing the parameter count, producing high-quality novel-view renderings on synthetic and real datasets. Notably, C-NGP models all $8$ scenes from the Real-LLFF dataset together, with only a $2.2\%$ drop in PSNR compared to vanilla NeRF, which models each scene independently. Further, C-NGP allows multiple style edits in the same network.