# Quality Resilient Deep Neural Networks

arXiv: https://arxiv.org/abs/1703.08119 · HTML: https://mlsanity.com/paper/1703.08119/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Mar 2017.

## Built on by (most cited first, 12 of 12)

- [Test-Time Training with Masked Autoencoders](https://mlsanity.com/paper/2209.07522/) (2022)
- [Test-Time Visual In-Context Tuning](https://mlsanity.com/paper/2503.21777/) (2025)
- [Towards Corruption-Agnostic Robust Domain Adaptation](https://mlsanity.com/paper/2104.10376/) (2021)
- [Benchmarking Robustness of Deep Learning Classifiers Using Two-Factor Perturbation](https://mlsanity.com/paper/2103.03102/) (2021)
- [Improving Robustness of Learning-based Autonomous Steering Using Adversarial Images](https://mlsanity.com/paper/2102.13262/) (2021)
- [Picking Up Quantization Steps for Compressed Image Classification](https://mlsanity.com/paper/2304.10714/) (2023)
- [Transparency Distortion Robustness for SOTA Image Segmentation Tasks](https://mlsanity.com/paper/2405.12864/) (2024)
- [Wiggling Weights to Improve the Robustness of Classifiers](https://mlsanity.com/paper/2111.09779/) (2021)
- [Benchmarking Robustness of Deep Learning Classifiers Using Two-Factor Perturbation](https://mlsanity.com/paper/2203.01323/) (2022)
- [Counterbalancing Teacher: Regularizing Batch Normalized Models for Robustness](https://mlsanity.com/paper/2207.01548/) (2022)
- [Towards Robust Training Datasets for Machine Learning with Ontologies: A Case Study for Emergency Road Vehicle Detection](https://mlsanity.com/paper/2406.15268/) (2024)
- [Discerning the Chaos: Detecting Adversarial Perturbations while Disentangling Intentional from Unintentional Noises](https://mlsanity.com/paper/2409.19619/) (2024)

## Abstract (arXiv)

We study deep neural networks for classification of images with quality distortions. We first show that networks fine-tuned on distorted data greatly outperform the original networks when tested on distorted data. However, fine-tuned networks perform poorly on quality distortions that they have not been trained for. We propose a mixture of experts ensemble method that is robust to different types of distortions. The "experts" in our model are trained on a particular type of distortion. The output of the model is a weighted sum of the expert models, where the weights are determined by a separate gating network. The gating network is trained to predict optimal weights for a particular distortion type and level. During testing, the network is blind to the distortion level and type, yet can still assign appropriate weights to the expert models. We additionally investigate weight sharing methods for the mixture model and show that improved performance can be achieved with a large reduction in the number of unique network parameters.