# Self-Supervised Deep Depth Denoising

arXiv: https://arxiv.org/abs/1909.01193 · HTML: https://mlsanity.com/paper/1909.01193/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Sep 2019.

## Built on by (most cited first, 15 of 15)

- [DynaMo: In-Domain Dynamics Pretraining for Visuo-Motor Control](https://mlsanity.com/paper/2409.12192/) (2024)
- [Self-supervised Learning for Sonar Image Classification](https://mlsanity.com/paper/2204.09323/) (2022)
- [RGB-Only Reconstruction of Tabletop Scenes for Collision-Free Manipulator Control](https://mlsanity.com/paper/2210.11668/) (2022)
- [Deep 360^circ Optical Flow Estimation Based on Multi-Projection Fusion](https://mlsanity.com/paper/2208.00776/) (2022)
- [Decoupling Fine Detail and Global Geometry for Compressed Depth Map Super-Resolution](https://mlsanity.com/paper/2411.03239/) (2024)
- [Geometry-aware Two-scale PIFu Representation for Human Reconstruction](https://mlsanity.com/paper/2112.02082/) (2021)
- [Token Boosting for Robust Self-Supervised Visual Transformer Pre-training](https://mlsanity.com/paper/2304.04175/) (2023)
- [DCL: Differential Contrastive Learning for Geometry-Aware Depth Synthesis](https://mlsanity.com/paper/2107.13087/) (2021)
- [Self-Annotated 3D Geometric Learning for Smeared Points Removal](https://mlsanity.com/paper/2311.09029/) (2023)
- [Unpaired Depth Super-Resolution in the Wild](https://mlsanity.com/paper/2105.12038/) (2021)
- [Refinement of Monocular Depth Maps via Multi-View Differentiable Rendering](https://mlsanity.com/paper/2410.03861/) (2024)
- [Centimeter-Wave Free-Space Time-of-Flight Imaging](https://mlsanity.com/paper/2105.11606/) (2021)
- [Accurate Ground-Truth Depth Image Generation via Overfit Training of Point Cloud Registration using Local Frame Sets](https://mlsanity.com/paper/2207.07016/) (2022)
- [Multi-Object Self-Supervised Depth Denoising](https://mlsanity.com/paper/2305.05778/) (2023)
- [VoxDepth: Rectification of Depth Images on Edge Devices](https://mlsanity.com/paper/2407.15067/) (2024)

## Abstract (arXiv)

Depth perception is considered an invaluable source of information for various vision tasks. However, depth maps acquired using consumer-level sensors still suffer from non-negligible noise. This fact has recently motivated researchers to exploit traditional filters, as well as the deep learning paradigm, in order to suppress the aforementioned non-uniform noise, while preserving geometric details. Despite the effort, deep depth denoising is still an open challenge mainly due to the lack of clean data that could be used as ground truth. In this paper, we propose a fully convolutional deep autoencoder that learns to denoise depth maps, surpassing the lack of ground truth data. Specifically, the proposed autoencoder exploits multiple views of the same scene from different points of view in order to learn to suppress noise in a self-supervised end-to-end manner using depth and color information during training, yet only depth during inference. To enforce selfsupervision, we leverage a differentiable rendering technique to exploit photometric supervision, which is further regularized using geometric and surface priors. As the proposed approach relies on raw data acquisition, a large RGB-D corpus is collected using Intel RealSense sensors. Complementary to a quantitative evaluation, we demonstrate the effectiveness of the proposed self-supervised denoising approach on established 3D reconstruction applications. Code is avalable at https://github.com/VCL3D/DeepDepthDenoising