# DeepCorrect: Correcting DNN models against Image Distortions

arXiv: https://arxiv.org/abs/1705.02406 · HTML: https://mlsanity.com/paper/1705.02406/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 May 2017.

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

- [Early-exit deep neural networks for distorted images: providing an efficient edge offloading](https://mlsanity.com/paper/2108.09343/) (2021)
- [ISP Distillation](https://mlsanity.com/paper/2101.10203/) (2021)
- [AIREPAIR: A Repair Platform for Neural Networks](https://mlsanity.com/paper/2211.15387/) (2022)
- [Fidelity Estimation Improves Noisy-Image Classification With Pretrained Networks](https://mlsanity.com/paper/2106.00673/) (2021)
- [A Comprehensive Overview of Fish-Eye Camera Distortion Correction Methods](https://mlsanity.com/paper/2401.00442/) (2023)
- [CD-COCO: A Versatile Complex Distorted COCO Database for Scene-Context-Aware Computer Vision](https://mlsanity.com/paper/2311.06976/) (2023)
- [Collaboratively Learning Federated Models from Noisy Decentralized Data](https://mlsanity.com/paper/2409.02189/) (2024)
- [Clarifying Myths About the Relationship Between Shape Bias, Accuracy, and Robustness](https://mlsanity.com/paper/2406.05006/) (2024)
- [Modeling Generalized Specialist Approach To Train Quality Resilient Snapshot Ensemble](https://mlsanity.com/paper/2206.05853/) (2022)
- [Impact of Scaled Image on Robustness of Deep Neural Networks](https://mlsanity.com/paper/2209.02132/) (2022)
- [Impact of Colour Variation on Robustness of Deep Neural Networks](https://mlsanity.com/paper/2209.02832/) (2022)
- [Impact of Light and Shadow on Robustness of Deep Neural Networks](https://mlsanity.com/paper/2305.14165/) (2023)
- [QNNRepair: Quantized Neural Network Repair](https://mlsanity.com/paper/2306.13793/) (2023)
- [PRUNE: A Patching Based Repair Framework for Certifiable Unlearning of Neural Networks](https://mlsanity.com/paper/2505.06520/) (2025)
- [Vision-Language Models Integrate Conflicting Evidence Like Biased Bayesian Observers](https://mlsanity.com/paper/2610.19407/) (2026)

## Abstract (arXiv)

In recent years, the widespread use of deep neural networks (DNNs) has facilitated great improvements in performance for computer vision tasks like image classification and object recognition. In most realistic computer vision applications, an input image undergoes some form of image distortion such as blur and additive noise during image acquisition or transmission. Deep networks trained on pristine images perform poorly when tested on such distortions. In this paper, we evaluate the effect of image distortions like Gaussian blur and additive noise on the activations of pre-trained convolutional filters. We propose a metric to identify the most noise susceptible convolutional filters and rank them in order of the highest gain in classification accuracy upon correction. In our proposed approach called DeepCorrect, we apply small stacks of convolutional layers with residual connections, at the output of these ranked filters and train them to correct the worst distortion affected filter activations, whilst leaving the rest of the pre-trained filter outputs in the network unchanged. Performance results show that applying DeepCorrect models for common vision tasks like image classification (ImageNet), object recognition (Caltech-101, Caltech-256) and scene classification (SUN-397), significantly improves the robustness of DNNs against distorted images and outperforms other alternative approaches..