Compression Artifacts Reduction by a Deep Convolutional Network

By Chao Dong, Yubin Deng, Chen Change Loy, Xiaoou Tang. IEEE International Conference on Computer Vision, 2015

874 citations recorded by the source corpus.

Lossy compression introduces complex compression artifacts, particularly the blocking artifacts, ringing effects and blurring. Existing algorithms either focus on removing blocking artifacts and produce blurred output, or restores sharpened images that are accompanied with ringing effects. Inspired by the deep convolutional networks (DCN) on super-resolution, we formulate a compact and efficient network for seamless attenuation of different compression artifacts. We also demonstrate that a deeper model can be effectively trained with the features learned in a shallow network. Following a similar "easy to hard" idea, we systematically investigate several practical transfer settings and show the effectiveness of transfer learning in low level vision problems. Our method shows superior performance than the state-of-the-arts both on the benchmark datasets and the real-world use cases (i.e. Twitter).

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/09642681d46282e76fd9d1336001ef6473b72ec8

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

Citations in the corpus

MemNet: A Persistent Memory Network for Image Restoration (2017, methodology): “As three classical image restoration tasks, image denoising aims to recover a clean image from a noisy observation, which commonly assumes additive white Gaussian noise with a standard deviation σ; single-image superresolution recovers a high-resolution (HR) image from a low-resolution (LR) image; a”
Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising (2016, methodology): “• For JPEG image deblocking, our DnCNN-3 is compared with two state-of-the-art methods, i.e., AR-CNN [41] and TNRD [16].”
Deep Learning Microscopy: Enhancing Resolution, Field-of-View and Depth-of-Field of Optical Microscopy Images Using Neural Networks (2017, background): “CNNs form a rapidly growing research field with various applications in e.g., image classification 3 , annotation 4 , style transfer 5 , compression 6 , and deconvolution in photography 7–9 , among others 10–13 .”
SwinIR: Image Restoration Using Swin Transformer (2021, methodology): “Dataset q ARCNN [17] DnCNN-3 [90] QGAC [20] RNAN [96] RDN [98] DRUNet [88] SwinIR (ours)”
MambaIR: A Simple Baseline for Image Restoration with State-Space Model (2024, background): “Image restoration has been significantly advanced since the introduction of deep learning by several pioneering works, such as SRCNN [16] for image super-4 resolution, DnCNN [81] for image denoising, ARCNN [15] for JPEG compression artifact reduction, etc .”
Residual Non-local Attention Networks for Image Restoration (2019, methodology): “…of modelling various vision problems, ranging from low-level (e.g., image denoising (Zhang et al., 2017a), compression artifacts reduction (Dong et al., 2015), and image super-resolution (Kim et al., 2016; Lai et al., 2017; Tai et al., 2017; Lim et al., 2017; Zhang et al., 2018a; Haris et…”
Deep Convolutional Neural Network for Inverse Problems in Imaging (2016, background): “Moreover, regression-type neural networks have demonstrated impressive results on inverse problems with exact models such as signal denoising [21], [22], deconvolution [23], artifact reduction [24], [25], signal recovery (modelbased restoration) [26], [27], and interpolation [28]–[30].”
Image Aesthetic Assessment: An experimental survey (2016, methodology): “Fine-tuning from trained CNN has been proved in [61], [62] to be an effective initialization approach.”
Learning Convolutional Networks for Content-Weighted Image Compression (2017, methodology): “Several traditional methods [20] and deep CNN models [6] have been proposed to tackle this issue.”
Video Frame Interpolation via Adaptive Separable Convolution (2017, methodology): “Our research is inspired by the success of applying deep learning to optical ﬂow estimation [2, 12, 16, 19, 50, 51, 52], artistic style transfer [17, 22, 28], and image enhancement [6, 9, 10, 46, 47, 55, 57, 62, 65].”
Multi-frame Quality Enhancement for Compressed Video (2018, methodology): “3Note that AR-CNN [9], DnCNN [40] and Li et al.”
Image Restoration Using Convolutional Auto-encoders with Symmetric Skip Connections (2016, neutral): “7 5.2 Image denoising . . . . . . . . . . . . . . 9 5.3 Image super-resolution . . . . . . . . . . 9 5.4 JPEG deblocking . . . . . . . . . . . . . 12 5.5 Image inpainting . . . . . . . . . . . . . 12”
Enhancing Quality for HEVC Compressed Videos (2017, methodology): “Unfortunately, the cost of visual quality enhancement is increased computational time [9], [16], particularly for CNN-based approaches.”
Palette: Image-to-Image Diffusion Models (2021, methodology): “[Dong et al. 2015] applied deep CNN architectures for JPEG restoration, and [Galteri et al.”
Deep Network Interpolation for Continuous Imagery Effect Transition (2018, background): “Deep convolutional neural (CNN) network has achieved a great success in many low-level vision tasks, such as image restoration [3, 18, 20, 1, 44, 2, 28], image-to-image translation [15, 37, 26, 47] and image style transfer [8, 17, 6].”