On classification of distorted images with deep convolutional neural networks

By Yiren Zhou, Sibo Song, Ngai-Man Cheung. 2017

121 citations recorded by the source corpus.

Image blur and image noise are common distortions during image acquisition. In this paper, we systematically study the effect of image distortions on the deep neural network (DNN) image classifiers. First, we examine the DNN classifier performance under four types of distortions. Second, we propose two approaches to alleviate the effect of image distortion: re-training and fine-tuning with noisy images. Our results suggest that, under certain conditions, fine-tuning with noisy images can alleviate much effect due to distorted inputs, and is more practical than re-training.

Subfields: Computer Vision and Pattern Recognition, Media Technology, Optimization & Training

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

In Optimization & Training

Citations in the corpus

On Robustness and Transferability of Convolutional Neural Networks (2020, background): “Subsequent studies have investigated brittleness to similar corruptions [53, 76], as well as to impulse noise [31], photometric perturbations [62], and small shifts and other transformations [2, 17, 74].”
A Study and Comparison of Human and Deep Learning Recognition Performance under Visual Distortions (2017, background): “Following the results of [13] and [14], we additionally finetune the networks on distorted data.”
Generalisation in humans and deep neural networks (2018, methodology): “[61]), or are today’s training methods / optimisers not sufficient to solve robust and general object recognition? In order to shed light on the dissimilarities we found, we performed a second batch of experiments by training networks directly on distorted images.”
CURE-OR: Challenging Unreal and Real Environments for Object Recognition (2018, background): “[17] also studied the effect of blur and noise in deep network-based recognition and showed their vulnerabilities.”
Human and DNN Classification Performance on Images With Quality Distortions (2019, methodology): “We additionally perform fine-tuning of the models with distorted training data (as in Vasiljevic et al. (2016) and Zhou et al. (2017)).”
Quality Resilient Deep Neural Networks (2017, methodology): “The method assumes knowledge of camera noise and blurring parameters, but for general applications (e.g., images from the Internet) the camera parameters may not be available.”
Biologically inspired sleep algorithm for increased generalization and adversarial robustness in deep neural networks (2020, background): “Likewise, ANNs trained on unperturbed images exhibit reduced performance when images in the test set are distorted, for example, through horizontal translations, blurring, or the addition of compression artifacts (Dodge & Karam (2016); Vasiljevic et al. (2016); Zhou et al. (2017)).”