mlsanityData to 7 Oct 2026

Aberration-Aware Depth-from-Focus

arXiv 2303.04654 ↗First posted 8 Mar 2023cs.CV eess.IV physics.opticscited by 12

Not on the map. Why?

Abstract (arXiv)

Computer vision methods for depth estimation usually use simple camera models with idealized optics. For modern machine learning approaches, this creates an issue when attempting to train deep networks with simulated data, especially for focus-sensitive tasks like Depth-from-Focus. In this work, we investigate the domain gap caused by off-axis aberrations that will affect the decision of the best-focused frame in a focal stack. We then explore bridging this domain gap through aberration-aware training (AAT). Our approach involves a lightweight network that models lens aberrations at different positions and focus distances, which is then integrated into the conventional network training pipeline. We evaluate the generality of pretrained models on both synthetic and real-world data. Our experimental results demonstrate that the proposed AAT scheme can improve depth estimation accuracy without fine-tuning the model or modifying the network architecture.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Builds on 30 of its 61 references that are in the map. First posted on arXiv 8 Mar 2023.

Builds on

30 of 61 references.

References · 30

  1. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 4,993
  2. 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,829
  3. 2022DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detectioncited by 1,306
  4. 2015A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimationcited by 878
  5. 2016Unsupervised Monocular Depth Estimation with Left-Right Consistencycited by 819
  6. 2020AdaBins: Depth Estimation using Adaptive Binscited by 596
  7. 2018Pyramid Stereo Matching Networkcited by 486
  8. 2018DeepMVS: Learning Multi-view Stereopsiscited by 277
  9. 2016DeMoN: Depth and Motion Network for Learning Monocular Stereocited by 200
  10. 2018BA-Net: Dense Bundle Adjustment Networkcited by 173
  11. 2021CFNet: Cascade and Fused Cost Volume for Robust Stereo Matchingcited by 153
  12. 2017CNN-SLAM: Real-time dense monocular SLAM with learned depth predictioncited by 152
18 more
  1. 2018CodeSLAM - Learning a Compact, Optimisable Representation for Dense Visual SLAMcited by 151
  2. 20193D Ken Burns Effect from a Single Imagecited by 128
  3. 2022Revealing the Dark Secrets of Masked Image Modelingcited by 117
  4. 2018Deep Depth Completion of a Single RGB-D Imagecited by 110
  5. 2022P3Depth: Monocular Depth Estimation with a Piecewise Planarity Priorcited by 106
  6. 2019CAM-Convs: Camera-Aware Multi-Scale Convolutions for Single-View Depthcited by 75
  7. 2019Deep Optics for Monocular Depth Estimation and 3D Object Detectioncited by 65
  8. 2020Single Image Depth Estimation Trained via Depth from Defocus Cuescited by 45
  9. 2020Single-shot Hyperspectral-Depth Imaging with Learned Diffractive Opticscited by 39
  10. 2020Focus on defocus: bridging the synthetic to real domain gap for depth estimationcited by 37
  11. 2023Curriculum Learning for ab initio Deep Learned Refractive Opticscited by 32
  12. 2017Deep Depth From Focuscited by 28
  13. 2021Bridging Unsupervised and Supervised Depth from Focus via All-in-Focus Supervisioncited by 23
  14. 2017Aperture Supervision for Monocular Depth Estimationcited by 21
  15. 2021Deep Depth from Focus with Differential Focus Volumecited by 20
  16. 2020Miniscope3D: optimized single-shot miniature 3D fluorescence microscopycited by 17
  17. 2018Deep Depth from Defocus: how can defocus blur improve 3D estimation using dense neural networks?cited by 16
  18. 2022Learning Depth from Focus in the Wildcited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 4; 2025: 6; 2026: 2.

12 papers.

  1. 2024PhoCoLens: Photorealistic and Consistent Reconstruction in Lensless Imagingcited by 13
  2. 2024A Differentiable Wave Optics Model for End-to-End Computational Imaging System Optimizationcited by 6
  3. 2024Limitations of Data-Driven Spectral Reconstruction -- An Optics-Aware Analysiscited by 5
  4. 2025Efficient Depth- and Spatially-Varying Image Simulation for Defocus Deblurcited by 4
  5. 2024Self-Supervised Spatially Variant PSF Estimation for Aberration-Aware Depth-from-Defocuscited by 2
  6. 2025Digitally Prototype Your Eye Tracker: Simulating Hardware Performance using 3D Synthetic Datacited by 2
  7. 2025Learning Latent Transmission and Glare Maps for Lens Veiling Glare Removalcited by 2
  8. 2025Simulating Dual-Pixel Images From Ray Tracing For Depth Estimationcited by 1
  9. 2025Towards Blind Lens Aberration Correction via Large LensLib Pre-training and Discrete Degradation Priorscited by 1
  10. 2025Generative Multi-Focus Image Fusioncited by 0
  11. 2026Zero-Shot Depth from Defocuscited by 0
  12. 2026Towards Minimal Focal Stack in Shape from Focuscited by 0