A GAN-based Tunable Image Compression System

Wu, Lirong; Huang, Kejie; Shen, Haibin

doi:10.1109/wacv45572.2020.9093387

Cited by 35 publications

(13 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Image Compression. There are several image compression methods based on deep learning, such as RNN-based networks [3], CNN-based networks [7] and generative adversarial networks (GAN) [4], which use pixel-level difference as distortion and do not consider downstream tasks. In addition, some advanced works took content information [8] and task information [9] [5] into consideration.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Semantic-assisted image compression

Sun¹,

Guo²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Conventional image compression methods typically aim at pixel-level consistency while ignoring the performance of downstream AI tasks. To solve this problem, this paper proposes a Semantic-Assisted Image Compression method (SAIC), which can maintain semantic-level consistency to enable high performance of downstream AI tasks. To this end, we train the compression network using semantic-level loss function. In particular, semantic-level loss is measured using gradient-based semantic weights mechanism (GSW). GSW directly consider downstream AI tasks' perceptual results. Then, this paper proposes a semantic-level distortion evaluation metric to quantify the amount of semantic information retained during the compression process. Experimental results show that the proposed SAIC method can retain more semantic-level information and achieve better performance of downstream AI tasks compared to the traditional deep learning-based method and the advanced perceptual method at the same compression ratio.

show abstract

Section: Related Workmentioning

confidence: 99%

“…In contrast, deep learning-based compression methods are powerful due to the joint optimization of the entire compression model and excellent learning ability. Deep learning-based image compression have been explored by convolutional autoencoder (CAE) [2], recurrent network (RNN) [3], and generative adversarial networks (GAN) [4].…”

Section: Introductionmentioning

confidence: 99%

Semantic-assisted image compression

Sun¹,

Guo²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The decoder is treated as the generator in a standard GAN, and the reconstructions are fed into a discriminator alongside real examples. [1,23] show that data lost during compression can be synthesised and the model can still generate visually pleasing results; the balance between reconstruction and generative performance can also vary [22].…”

Section: State-of-the-art Lossy Compression Schemesmentioning

confidence: 99%

TACTIC: Joint Rate-Distortion-Accuracy Optimisation for Low Bitrate Compression

Kubiak,

Hadfield

2021

Preprint

View full text Add to dashboard Cite

We present TACTIC: Task-Aware Compression Through Intelligent Coding. Our lossy compression model learns based on the rate-distortion-accuracy trade-off for a specific task. By considering what information is important for the follow-on problem, the system trades off visual fidelity for good task performance at a low bitrate. When compared against JPEG at the same bitrate, our approach is able to improve the accuracy of ImageNet subset classification by 4.5%. We also demonstrate the applicability of our approach to other problems, providing a 3.4% accuracy and 4.9% mean IoU improvements in performance over task-agnostic compression for semantic segmentation.

show abstract

“…However, [7] adopts a complex and heuristic training procedure for stable training, and it is a non-standard GAN training. [6], [8] achieved to reconstruct realistic images at low bitrates. These methods use no rate term in their loss functions for stable training, but it can lead to suboptimal bitrates.…”

Section: B Learned Image Compressionmentioning

confidence: 99%

Fidelity-Controllable Extreme Image Compression with Generative Adversarial Networks

Iwai

Miyazaki

Sugaya

et al. 2020

Preprint

View full text Add to dashboard Cite

We propose a GAN-based image compression method working at extremely low bitrates below 0.1bpp. Most existing learned image compression methods suffer from blur at extremely low bitrates. Although GAN can help to reconstruct sharp images, there are two drawbacks. First, GAN makes training unstable. Second, the reconstructions often contain unpleasing noise or artifacts. To address both of the drawbacks, our method adopts two-stage training and network interpolation. The twostage training is effective to stabilize the training. Moreover, the network interpolation utilizes the models in both stages and reduces undesirable noise and artifacts, while maintaining important edges. Hence, we can control the trade-off between perceptual quality and fidelity without re-training models. The experimental results show that our model can reconstruct high quality images. Furthermore, our user study confirms that our reconstructions are preferable to state-of-the-art GAN-based image compression model. Our code will be available at https: //github.com/iwa-shi/fidelity controllable compression

show abstract

A GAN-based Tunable Image Compression System

Cited by 35 publications

References 9 publications

Semantic-assisted image compression

Semantic-assisted image compression

TACTIC: Joint Rate-Distortion-Accuracy Optimisation for Low Bitrate Compression

Fidelity-Controllable Extreme Image Compression with Generative Adversarial Networks

Contact Info

Product

Resources

About