Files and links (1)
CC BY V4.0, Open Access
Conference proceeding
Normal-guided Detail-Preserving Neural Implicit Function for High-Fidelity 3D Surface Reconstruction
Proceedings of the ACM on computer graphics and interactive techniques, Vol.8(1), 12
22/05/2025
Abstract
Neural implicit representations have emerged as a powerful paradigm for 3D reconstruction. However, despite their success, existing methods fail to capture fine geometric details and thin structures, especially in scenarios where only sparse multi-view RGB images of the objects of interest are available. This paper shows that training neural representations with first-order differential properties (surface normals) leads to highly accurate 3D surface reconstruction, even with as few as two RGB images. Using input RGB images, we compute approximate ground-truth surface normals from depth maps produced by an off-the-shelf monocular depth estimator. During training, we directly locate the surface point of the SDF network and supervise its normal with the one estimated from the depth map. Extensive experiments demonstrate that our method achieves state-of-the-art reconstruction accuracy with a minimal number of views, capturing intricate geometric details and thin structures that were previously challenging to capture. The source code and additional results are available at https://graphics-research-group.github.io/sn-nir.
Details
- Title
- Normal-guided Detail-Preserving Neural Implicit Function for High-Fidelity 3D Surface Reconstruction
- Authors/Creators
- Aarya Patel - Indraprastha Institute of Information Technology DelhiHamid Laga - Murdoch University, Centre for Biosecurity and One HealthOjaswa Sharma - Indraprastha Institute of Information Technology Delhi
- Publication Details
- Proceedings of the ACM on computer graphics and interactive techniques, Vol.8(1), 12
- Publisher
- ACM
- Number of pages
- 24
- Identifiers
- 991005788082807891
- Copyright
- © 2025 Copyright held by the owner/author(s).
- Murdoch Affiliation
- Centre for Biosecurity and One Health; School of Information Technology; Centre for Healthy Ageing
- Language
- English
- Resource Type
- Conference proceeding
Metrics
1 File views/ downloads
24 Record Views