Deep Generative Adversarial Compression Artifact Removal
Удаление артефактов сжатия с помощью глубокой генеративно-состязательной модели
2017-10-01
SCID: 54.1/pyf7uvke
Discuss with AI
compression artifact removalfully convolutional residual networkgenerative adversarial networkobject detectionstructural similarity (SSIM)
Figures from the paper
Abstract (AI)
Compression artifacts arise in images whenever a lossy compression algorithm is applied. These artifacts eliminate details present in the original image, or add noise and small structures; because of these effects they make images less pleasant for the human eye, and may also lead to decreased performance of computer vision algorithms such as object detectors. To eliminate such artifacts, when decompressing an image, it is required to recover the original image from a disturbed version. To this end, we present a feed-forward fully convolutional residual network model trained using a generative adversarial framework. To provide a baseline, we show that our model can be also trained optimizing the Structural Similarity (SSIM), which is a better loss with respect to the simpler Mean Squared Error (MSE). Our GAN is able to produce images with more photorealistic details than MSE or SSIM based networks. Moreover we show that our approach can be used as a pre-processing step for object detection in case images are degraded by compression to a point that state-of-the art detectors fail. In this task, our GAN method obtains better performance than MSE or SSIM trained networks.
Key Findings
1
A generative adversarial formulation built on the same architecture produces more photorealistic details than MSE- or SSIM-trained networks.
2
Directly optimizing Structural Similarity (SSIM) provides a more suitable artifact-removal objective than Mean Squared Error (MSE).
3
For compression-degraded object detection, the GAN-based method outperforms networks trained with MSE or SSIM.
4
The paper presents a fully convolutional residual network for removing image compression artifacts during decompression.
5
The proposed GAN improves object-detection performance on heavily compressed images where state-of-the-art detectors otherwise fail.
Research Object
images degraded by lossy compression
Research Subject
recovery of original image details and photorealistic structures, including the impact on object-detection performance
Publication Details
Publication Date
2017-10-01
Journal
Publisher
ISSN
Cited by
222
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai7
Deep Residual Learning for Image Recognition2016
Very Deep Convolutional Networks for Large-Scale Image Recognition2014
Adam: A Method for Stochastic Optimization2014
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network2017
Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising2017
Image Style Transfer Using Convolutional Neural Networks2016
Data Driven VRP: A Neural Network Model to Learn Hidden Preferences for VRP2021