Abstract
Most of existing compression artifacts reduction methods focused on the application for low-quality images and usually assumed a known compression quality factor. However, images compressed with high quality should also be manipulated because even small artifacts become noticeable when we enhance the compressed image. Also, the use of quality factor from the decoder is not practical because there are too many recompressed or transcoded images whose quality factor are not reliable and spatially varying. To address these issues, we propose a quality-adaptive artifacts removal network based on the gating scheme, with a quality estimator that works for a wide range of quality factor. Specifically, the estimator gives a pixel-wise quality factor, and our gating scheme generates gate-weights from the quality factor. Then, the gate-weights control the magnitudes of feature maps in our artifacts removal network. Thus, our gating scheme guarantees the proposed network to perform adaptively without changing the parameters according to the change of quality factor. Moreover, we exploit the Discrete Cosine Transform (DCT) scheme with 3D convolution for capturing both spatial and frequency dependencies of images. Experiments show that the proposed network provides better performance than the state-of-the-art methods over a wide range of quality factor. Also, the proposed method provides robust results in real-world scenarios such as the manipulation of transcoded images and videos.
Highlights
There have been many deep convolutional neural networks (CNNs) [1]–[3] for reducing various kinds of noise in images and videos [4]–[12]
As mentioned in the introduction, we think that both schemes have certain limitations, and we propose a flexible network named as adaptively gated artifact removal network (AGARNet)
We find that processing in the dual-domain with the Inverse Discrete Cosine Transform (IDCT) loss and 3D convolution provides the best architecture for compression artifacts removal
Summary
There have been many deep convolutional neural networks (CNNs) [1]–[3] for reducing various kinds of noise in images and videos [4]–[12]. Non-blind methods have a strong disadvantage that they need multiple networks that are trained to different noise levels to cope with various compression quality factors. They generally require large memory, and there can be redundancy between the models for similar Qs. we need to consider the properties of real-world compressed images for their enhancement and denoising. We design a Q-estimator, which finds the spatially variant quality-level map for the given input This is essential for dealing with compressed images and videos in the real world. The proposed method is shown to provide better performance than the conventional ones, and works well for the recompressed images
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
