A Radiance Field Loss for Fast and Simple Emissive Surface Reconstruction

2025-01-27

AI Summary

Implicit radiance field methods for glossy surface reconstruction suffer from high computational cost and complex explicit geometry extraction. Method: We propose an efficient, emission-aware reconstruction framework that abandons conventional alpha compositing and ray marching in volumetric rendering. Instead, we introduce a novel projection-supervised radiance field loss function that directly constrains radiance values on the 2D image plane, endowing them with explicit physical semantics. Coupled with level-set-based implicit surface extraction and an enhanced Instant NGP architecture, our method enables end-to-end explicit emissive surface reconstruction. Results: Experiments demonstrate near real-time inference speed while incurring only a marginal PSNR drop of 0.1 dB. Our approach significantly simplifies the pipeline, avoids exponential voxel-based representations, and outputs high-fidelity, editable explicit geometry and material properties.

Technology Category

• Computer Vision: Computational Photography, Image & Video Synthesis
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

We present a fast and simple technique to convert images into an emissive surface-based scene representation. Building on existing emissive volume reconstruction algorithms, we introduce a subtle yet impactful modification of the loss function requiring changes to only a few lines of code: instead of integrating the radiance field along rays and supervising the resulting images, we project the training images into the scene to directly supervise the spatio-directional radiance field. The primary outcome of this change is the complete removal of alpha blending and ray marching from the image formation model, instead moving these steps into the loss computation. In addition to promoting convergence to surfaces, this formulation assigns explicit semantic meaning to 2D subsets of the radiance field, turning them into well-defined emissive surfaces. We finally extract a level set from this representation, which results in a high-quality emissive surface model. Our method retains much of the speed and quality of the baseline algorithm. For instance, a suitably modified variant of Instant~NGP maintains comparable computational efficiency, while achieving an average PSNR that is only 0.1 dB lower. Most importantly, our method generates explicit surfaces in place of an exponential volume, doing so with a level of simplicity not seen in prior work.

Problem

• Real-time Rendering
• Illuminated Surfaces
• Efficient Image Processing

Innovation

• Simplified Radiant Surface Reconstruction
• Direct Illumination Guidance
• Efficient Rendering Technique