# Depth Estimators Are Implicit Neural Fields for 3D Scene Geometry Inpainting and Reconstruction

arXiv: https://arxiv.org/abs/2607.16286 · HTML: https://mlsanity.com/paper/2607.16286/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 44 references that are in the map.
- First posted on arXiv 11 Jul 2026.

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

- [Depth Anything 3: Recovering the Visual Space from Any Views](https://mlsanity.com/paper/2511.10647/) (2025)
- [Virtual KITTI 2](https://mlsanity.com/paper/2001.10773/) (2020)
- [Depth-Regularized Optimization for 3D Gaussian Splatting in Few-Shot Images](https://mlsanity.com/paper/2311.13398/) (2023)
- [InFusion: Inpainting 3D Gaussians via Learning Depth Completion from Diffusion Prior](https://mlsanity.com/paper/2404.11613/) (2024)
- [3D Gaussian Inpainting with Depth-Guided Cross-View Consistency](https://mlsanity.com/paper/2502.11801/) (2025)
- [InfiniDepth: Arbitrary-Resolution and Fine-Grained Depth Estimation with Neural Implicit Fields](https://mlsanity.com/paper/2601.03252/) (2026)
- [Skyfall-GS: Synthesizing Immersive 3D Urban Scenes from Satellite Imagery](https://mlsanity.com/paper/2510.15869/) (2025)
- [SkySplat: Generalizable 3D Gaussian Splatting from Multi-Temporal Sparse Satellite Images](https://mlsanity.com/paper/2508.09479/) (2025)
- [SplatFill: 3D Scene Inpainting via Depth-Guided Gaussian Splatting](https://mlsanity.com/paper/2509.07809/) (2025)
- [Endowing GPT-4 with a Humanoid Body: Building the Bridge Between Off-the-Shelf VLMs and the Physical World](https://mlsanity.com/paper/2511.00041/) (2025)

## Abstract (arXiv)

The 3D geometry of real-world scene data is often incomplete. Mainstream methods use depth estimators to inpaint missing structure. However, their prediction results can be inconsistent with observed geometry, or unreliable on out-of-distribution data. To solve these problems, we propose Neural Depth Field (NDF). Our key insight is that a depth estimator can also be a scene-level implicit field. As an estimator, it adapts to the target domain by learning observed depth data. As an implicit field, it fits the existing geometry to maintain consistency. Under this view, NDF addresses both problems through a single test-time optimization. Experiments show that NDF produces high-fidelity and globally consistent geometry across diverse scene data, ranging from indoor scans to satellite imagery. It reduces cross-view inconsistency by 63.3% and improves inpainting accuracy by 23.1%, achieving state-of-the-art performance in 3D scene geometry inpainting. The code is available at: https://github.com/Shadow-Dream/Neural-Depth-Field.