# Examining the Impact of Blur on Recognition by Convolutional Networks

arXiv: https://arxiv.org/abs/1611.05760 · HTML: https://mlsanity.com/paper/1611.05760/

## Key facts

- Cited by 58 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Nov 2016.

## Built on by (most cited first, 30 of 58)

- [Test-Time Training with Masked Autoencoders](https://mlsanity.com/paper/2209.07522/) (2022)
- [A Study of Face Obfuscation in ImageNet](https://mlsanity.com/paper/2103.06191/) (2021)
- [Towards Robust Referring Image Segmentation](https://mlsanity.com/paper/2209.09554/) (2022)
- [When Human Pose Estimation Meets Robustness: Adversarial Algorithms and Benchmarks](https://mlsanity.com/paper/2105.06152/) (2021)
- [Improving robustness against common corruptions with frequency biased models](https://mlsanity.com/paper/2103.16241/) (2021)
- [Defending Against Image Corruptions Through Adversarial Augmentations](https://mlsanity.com/paper/2104.01086/) (2021)
- [Introspective Learning : A Two-Stage Approach for Inference in Neural Networks](https://mlsanity.com/paper/2209.08425/) (2022)
- [Robust Semantic Segmentation with Superpixel-Mix](https://mlsanity.com/paper/2108.00968/) (2021)
- [Joint Face Image Restoration and Frontalization for Recognition](https://mlsanity.com/paper/2105.09907/) (2021)
- [Improving Diversity with Adversarially Learned Transformations for Domain Generalization](https://mlsanity.com/paper/2206.07736/) (2022)
- [Augraphy: A Data Augmentation Library for Document Images](https://mlsanity.com/paper/2208.14558/) (2022)
- [Classification robustness to common optical aberrations](https://mlsanity.com/paper/2308.15499/) (2023)
- [Contrastive Reasoning in Neural Networks](https://mlsanity.com/paper/2103.12329/) (2021)
- [LDP-Feat: Image Features with Local Differential Privacy](https://mlsanity.com/paper/2308.11223/) (2023)
- [Noise Optimization for Artificial Neural Networks](https://mlsanity.com/paper/2102.04450/) (2021)
- [Deep Learning based Computer Vision Methods for Complex Traffic Environments Perception: A Review](https://mlsanity.com/paper/2211.05120/) (2022)
- [On the Robustness of Object Detection Models on Aerial Images](https://mlsanity.com/paper/2308.15378/) (2023)
- [Test-Time Visual In-Context Tuning](https://mlsanity.com/paper/2503.21777/) (2025)
- [AQuA: Analytical Quality Assessment for Optimizing Video Analytics Systems](https://mlsanity.com/paper/2101.09752/) (2021)
- [Exploring Resolution and Degradation Clues as Self-supervised Signal for Low Quality Object Detection](https://mlsanity.com/paper/2208.03062/) (2022)
- [Certifying Model Accuracy under Distribution Shifts](https://mlsanity.com/paper/2201.12440/) (2022)
- [The Causal Structure of Domain Invariant Supervised Representation Learning](https://mlsanity.com/paper/2208.06987/) (2022)
- [Does lossy image compression affect racial bias within face recognition?](https://mlsanity.com/paper/2208.07613/) (2022)
- [Improving Model Generalization by On-manifold Adversarial Augmentation in the Frequency Domain](https://mlsanity.com/paper/2302.14302/) (2023)
- [Towards Corruption-Agnostic Robust Domain Adaptation](https://mlsanity.com/paper/2104.10376/) (2021)
- [PoseBench: Benchmarking the Robustness of Pose Estimation Models under Corruptions](https://mlsanity.com/paper/2406.14367/) (2024)
- [Benchmarking Robustness of Deep Learning Classifiers Using Two-Factor Perturbation](https://mlsanity.com/paper/2103.03102/) (2021)
- [VITA: A Multi-Source Vicinal Transfer Augmentation Method for Out-of-Distribution Generalization](https://mlsanity.com/paper/2204.11531/) (2022)
- [Learning to See Through a Baby's Eyes: Early Visual Diets Enable Robust Visual Intelligence in Humans and Machines](https://mlsanity.com/paper/2511.14440/) (2025)
- [Improving Robustness of Learning-based Autonomous Steering Using Adversarial Images](https://mlsanity.com/paper/2102.13262/) (2021)

## Abstract (arXiv)

State-of-the-art algorithms for many semantic visual tasks are based on the use of convolutional neural networks. These networks are commonly trained, and evaluated, on large annotated datasets of artifact-free high-quality images. In this paper, we investigate the effect of one such artifact that is quite common in natural capture settings: optical blur. We show that standard network models, trained only on high-quality images, suffer a significant degradation in performance when applied to those degraded by blur due to defocus, or subject or camera motion. We investigate the extent to which this degradation is due to the mismatch between training and input image statistics. Specifically, we find that fine-tuning a pre-trained model with blurred images added to the training set allows it to regain much of the lost accuracy. We also show that there is a fair amount of generalization between different degrees and types of blur, which implies that a single network model can be used robustly for recognition when the nature of the blur in the input is unknown. We find that this robustness arises as a result of these models learning to generate blur invariant representations in their hidden layers. Our findings provide useful insights towards developing vision systems that can perform reliably on real world images affected by blur.