Feature Consistency Training With JPEG Compressed Images

Deep neural networks (DNNs) are recently found to be vulnerable to JPEG compression artifacts, which distort the feature representations of DNNs leading to serious accuracy degradation. Most existing training methods which aim to address this problem add compressed images to the training data to enh...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on circuits and systems for video technology Vol. 30; no. 12; pp. 4769 - 4780
Main Authors Wan, Sheng, Wu, Tung-Yu, Hsu, Heng-Wei, Wong, Wing Hung, Lee, Chen-Yi
Format Journal Article
LanguageEnglish
Published New York IEEE 01.12.2020
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Artificial neural networks
classification robustness
Compression artifacts
Consistency
deep neural network
Distortion
Feature extraction
Image coding
Image compression
Image enhancement
Image quality
Iterative methods
JPEG compression
JPEG encoders-decoders
License plates
Q factors
Quantization (signal)
Representations
Robustness
Training
Training data
Transform coding
Online AccessGet full text

Cover

More Information
Summary:Deep neural networks (DNNs) are recently found to be vulnerable to JPEG compression artifacts, which distort the feature representations of DNNs leading to serious accuracy degradation. Most existing training methods which aim to address this problem add compressed images to the training data to enhance the robustness of DNNs. However, their improvements are limited since these methods usually regard the compressed images as new training samples instead of distorted samples. The feature distortions between the raw images and the compressed images are not investigated. In this work, we propose a new training method, called Feature Consistency Training, that is designed to minimize the feature distortions caused by JPEG artifacts. At each training iteration, we simultaneously input a raw image and its compressed version with a randomly sampled quality into a DNN model and extract the features from the internal layers. By adding feature consistency constraint to the objective function, the feature distortions in the representation space are minimized in order to learn robust filters. Besides, we present a residual mapping block which takes the quality factor of the compressed image as an additional information to further reduce the feature distortion. Extensive experiments demonstrate that our method outperforms several existed training methods on JPEG compressed images. Furthermore, DNN models trained by our method are found to be more robust to unseen distortions.
ISSN:1051-8215
1558-2205
DOI:10.1109/TCSVT.2019.2959815