# Fast Light-Weight Near-Field Photometric Stereo

arXiv: https://arxiv.org/abs/2203.16515 · HTML: https://mlsanity.com/paper/2203.16515/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 35 references that are in the map.
- First posted on arXiv 30 Mar 2022.

## Builds on (most cited first, 10 of 10)

- [Self-calibrating Deep Photometric Stereo Networks](https://mlsanity.com/paper/1903.07366/) (2019)
- [CNN-PS: CNN-based Photometric Stereo for General Non-Convex Surfaces](https://mlsanity.com/paper/1808.10093/) (2018)
- [Shape and Material Capture at Home](https://mlsanity.com/paper/2104.06397/) (2021)
- [Uncalibrated Neural Inverse Rendering for Photometric Stereo of General Surfaces](https://mlsanity.com/paper/2012.06777/) (2020)
- [SilNet : Single- and Multi-View Reconstruction by Learning from Silhouettes](https://mlsanity.com/paper/1711.07888/) (2017)
- [Normal Integration: A Survey](https://mlsanity.com/paper/1709.05940/) (2017)
- [LUCES: A Dataset for Near-Field Point Light Source Photometric Stereo](https://mlsanity.com/paper/2104.13135/) (2021)
- [Variational Uncalibrated Photometric Stereo under General Lighting](https://mlsanity.com/paper/1904.03942/) (2019)
- [LED-based Photometric Stereo: Modeling, Calibration and Numerical Solution](https://mlsanity.com/paper/1707.01018/) (2017)
- [A CNN Based Approach for the Near-Field Photometric Stereo Problem](https://mlsanity.com/paper/2009.05792/) (2020)

## Built on by (most cited first, 12 of 12)

- [S3-NeRF: Neural Reflectance Field from Shading and Shadow under a Single Viewpoint](https://mlsanity.com/paper/2210.08936/) (2022)
- [MVPSNet: Fast Generalizable Multi-view Photometric Stereo](https://mlsanity.com/paper/2305.11167/) (2023)
- [Leveraging Near-Field Lighting for Monocular Depth Estimation from Endoscopy Videos](https://mlsanity.com/paper/2403.17915/) (2024)
- [Light of Normals: Unified Feature Representation for Universal Photometric Stereo](https://mlsanity.com/paper/2506.18882/) (2025)
- [A Surface-normal Based Neural Framework for Colonoscopy Reconstruction](https://mlsanity.com/paper/2303.07264/) (2023)
- [MERLiN: Single-Shot Material Estimation and Relighting for Photometric Stereo](https://mlsanity.com/paper/2409.00674/) (2024)
- [NFL-BA: Near-Field Light Bundle Adjustment for SLAM in Dynamic Lighting](https://mlsanity.com/paper/2412.13176/) (2024)
- [MS-PS: A Multi-Scale Network for Photometric Stereo With a New Comprehensive Training Dataset](https://mlsanity.com/paper/2211.14118/) (2022)
- [GaNI: Global and Near Field Illumination Aware Neural Inverse Rendering](https://mlsanity.com/paper/2403.15651/) (2024)
- [PPS-Ctrl: Controllable Sim-to-Real Translation for Colonoscopy Depth Estimation](https://mlsanity.com/paper/2504.17067/) (2025)
- [Near-Light Color Photometric Stereo for Mono-Chromatic Non-Lambertian Surfaces](https://mlsanity.com/paper/2601.12666/) (2026)
- [Near-light Photometric Stereo with Symmetric Lights](https://mlsanity.com/paper/2603.16404/) (2026)

## Abstract (arXiv)

We introduce the first end-to-end learning-based solution to near-field Photometric Stereo (PS), where the light sources are close to the object of interest. This setup is especially useful for reconstructing large immobile objects. Our method is fast, producing a mesh from 52 512×384 resolution images in about 1 second on a commodity GPU, thus potentially unlocking several AR/VR applications. Existing approaches rely on optimization coupled with a far-field PS network operating on pixels or small patches. Using optimization makes these approaches slow and memory intensive (requiring 17GB GPU and 27GB of CPU memory) while using only pixels or patches makes them highly susceptible to noise and calibration errors. To address these issues, we develop a recursive multi-resolution scheme to estimate surface normal and depth maps of the whole image at each step. The predicted depth map at each scale is then used to estimate `per-pixel lighting' for the next scale. This design makes our approach almost 45× faster and 2circ more accurate (11.3circ vs. 13.3circ Mean Angular Error) than the state-of-the-art near-field PS reconstruction technique, which uses iterative optimization.