# Learning Lens Blur Fields

arXiv: https://arxiv.org/abs/2310.11535 · HTML: https://mlsanity.com/paper/2310.11535/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 85 references that are in the map.
- First posted on arXiv 17 Oct 2023.

## Builds on (most cited first, 17 of 17)

- [Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains](https://mlsanity.com/paper/2006.10739/) (2020)
- [Neural Fields in Visual Computing and Beyond](https://mlsanity.com/paper/2111.11426/) (2021)
- [Learning a Convolutional Neural Network for Non-uniform Motion Blur Removal](https://mlsanity.com/paper/1503.00593/) (2015)
- [Neural Blind Deconvolution Using Deep Priors](https://mlsanity.com/paper/1908.02197/) (2019)
- [Synthetic Depth-of-Field with a Single-Camera Mobile Phone](https://mlsanity.com/paper/1806.04171/) (2018)
- [Deep Optics for Single-shot High-dynamic-range Imaging](https://mlsanity.com/paper/1908.00620/) (2019)
- [DiffuserCam: Lensless Single-exposure 3D Imaging](https://mlsanity.com/paper/1710.02134/) (2017)
- [Single-shot Hyperspectral-Depth Imaging with Learned Diffractive Optics](https://mlsanity.com/paper/2009.00463/) (2020)
- [Learning Fully Convolutional Networks for Iterative Non-blind Deconvolution](https://mlsanity.com/paper/1611.06495/) (2016)
- [Learning to Autofocus](https://mlsanity.com/paper/2004.12260/) (2020)
- [Defocus Map Estimation and Deblurring from a Single Dual-Pixel Image](https://mlsanity.com/paper/2110.05655/) (2021)
- [Learning a Discriminative Prior for Blind Image Deblurring](https://mlsanity.com/paper/1803.03363/) (2018)
- [Spatially-Variant CNN-based Point Spread Function Estimation for Blind Deconvolution and Depth Estimation in Optical Microscopy](https://mlsanity.com/paper/2010.04011/) (2020)
- [On Regularizing Coordinate-MLPs](https://mlsanity.com/paper/2202.00790/) (2022)
- [Deep Learning for space-variant deconvolution in galaxy surveys](https://mlsanity.com/paper/1911.00443/) (2019)
- [Multi-View Motion Synthesis via Applying Rotated Dual-Pixel Blur Kernels](https://mlsanity.com/paper/2111.07837/) (2021)
- [Full-field Fourier ptychography (FFP): spatially varying pupil modeling and its application for rapid field-dependent aberration metrology](https://mlsanity.com/paper/1905.00162/) (2019)

## Built on by (most cited first, 6 of 6)

- [A Physics-Informed Blur Learning Framework for Imaging Systems](https://mlsanity.com/paper/2502.11382/) (2025)
- [Representing Domain-Mixing Optical Degradation for Real-World Computational Aberration Correction via Vector Quantization](https://mlsanity.com/paper/2403.10012/) (2024)
- [ConVRT: Consistent Video Restoration Through Turbulence with Test-time Optimization of Neural Video Representations](https://mlsanity.com/paper/2312.04679/) (2023)
- [CircleFlow: Flow-Guided Camera Blur Estimation using a Circle Grid Target](https://mlsanity.com/paper/2512.00796/) (2025)
- [Face2Scene: Using Facial Degradation as an Oracle for Diffusion-Based Scene Restoration](https://mlsanity.com/paper/2603.16570/) (2026)
- [Broadband Wide Field of View Imaging with Computational Mirrors](https://mlsanity.com/paper/2605.00029/) (2026)

## Abstract (arXiv)

Optical blur is an inherent property of any lens system and is challenging to model in modern cameras because of their complex optical elements. To tackle this challenge, we introduce a high-dimensional neural representation of blur-textitthe lens blur field-and a practical method for acquiring it. The lens blur field is a multilayer perceptron (MLP) designed to (1) accurately capture variations of the lens 2D point spread function over image plane location, focus setting and, optionally, depth and (2) represent these variations parametrically as a single, sensor-specific function. The representation models the combined effects of defocus, diffraction, aberration, and accounts for sensor features such as pixel color filters and pixel-specific micro-lenses. To learn the real-world blur field of a given device, we formulate a generalized non-blind deconvolution problem that directly optimizes the MLP weights using a small set of focal stacks as the only input. We also provide a first-of-its-kind dataset of 5D blur fields-for smartphone cameras, camera bodies equipped with a variety of lenses, etc. Lastly, we show that acquired 5D blur fields are expressive and accurate enough to reveal, for the first time, differences in optical behavior of smartphone devices of the same make and model. Code and data can be found at blur-fields.github.io.