DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better

By Orest Kupyn, T. Martyniuk, Junru Wu, Zhangyang Wang. 2019

1,132 citations recorded by the source corpus.

We present a new end-to-end generative adversarial network (GAN) for single image motion deblurring, named DeblurGAN-V2, which considerably boosts state-of-the-art deblurring performance while being much more flexible and efficient. DeblurGAN-V2 is based on a relativistic conditional GAN with a double-scale discriminator. For the first time, we introduce the Feature Pyramid Network into deblurring, as a core building block in the generator of DeblurGAN-V2. It can flexibly work with a wide range of backbones, to navigate the balance between performance and efficiency. The plug-in of sophisticated backbones (e.g. Inception ResNet v2) can lead to solid state-of-the-art performance. Meanwhile, with light-weight backbones (e.g. MobileNet and its variants), DeblurGAN-V2 becomes 10-100 times faster than the nearest competitors, while maintaining close to state-of-the-art results, implying the option of real-time video deblurring. We demonstrate that DeblurGAN-V2 has very competitive performance on several popular benchmarks, in terms of deblurring quality (both objective and subjective), as well as efficiency. In addition, we show the architecture to be effective for general image restoration tasks too. Our models and codes will be made available upon acceptance.

Subfields: Computer Vision and Pattern Recognition

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

AutoGAN-Distiller: Searching to Compress Generative Adversarial Networks (2020, neutral): “These encoder-decoder GANs also gain popularity in image enhancement (Wang et al., 2018a; Kupyn et al., 2019; Jiang et al., 2019) and editing (Sanakoyeu et al.”
Pre-Trained Image Processing Transformer (2020, methodology): “Method MSCNN [54] SRN [67] DSD [30] DeblurGANv2 [44] DMPHN [84] LEBMD [40] EDSD [81]”
TransGAN: Two Pure Transformers Can Make One Strong GAN, and That Can Scale Up (2021, background): “GANs are widely adopted in image translation [3, 4, 41], image enhancement [7, 42, 43], and image editing [44, 45].”
Image-to-Image Translation: Methods and Applications (2021, background): “Image deraining [195], [196], [197], image dehazing [199], [200], [201], [198], [202] and image deblurring [206], [203], [204], [205] aim to remove optical distortions from photos that were taken out of focus or while the camera was moving, or from photos of faraway geographical or astronomical feat”
Blind Face Restoration via Deep Multi-scale Component Dictionaries (2020, background): “, image super-resolution [9,36,46], denoising [42,43], and deblurring [22, 23], blind image restoration suffers from more challenges, yet is of great practical value in restoring real LQ images.”
Real-World Blur Dataset for Learning and Benchmarking Deblurring Algorithms (2020, methodology): “(e) DeblurGAN [21] (f) DeblurGAN-v2 [22]”
Uformer: A General U-Shaped Transformer for Image Restoration (2021, methodology): “[67], DeblurGAN-v2 [29], DBGAN [80], SPAIR [47], Zhang et al.”
MAXIM: Multi-Axis MLP for Image Processing (2022, background): “Residual and dense connections [42, 93, 93, 117, 118], the multi-scale feature learning [19,40,95], attention mechanisms [64,89,107,108,118], and non-local networks [52, 91, 118] are such good examples.”
Advancing Image Understanding in Poor Visibility Environments: A Collective Benchmark Study (2019, background): “There are numerous algorithms aiming to enhance visibility of the degraded imagery, such as image and video denoising/inpainting [74]–[78], deblurring [79], [80], [159], [163], [164] super-resolution [81]–[84] and interpolation [85].”
Multi-Stage Progressive Image Restoration (2021, methodology): “Reference Blurry SRN [69] DeblurGANv2 [41] Gao et al.”
Learning Enriched Features for Fast Image Restoration and Enhancement (2022, methodology): “The encoder-decoder models [1], [2], [3], [4] first progressively map the input to a lowresolution representation, and then apply a gradual reverse mapping to the original resolution.”
Deblurring by Realistic Blurring (2020, background): “Kupyn et al. propose DeblurGAN [16] and DeblurGAN-v2 [17] to remove blur kernels based on adversarial learning.”
Towards Real-World Blind Face Restoration with Generative Facial Prior (2021, methodology): “We also compare our GFP-GAN with image restoration methods: RCAN [75], ESRGAN [66] and DeblurGANv2 [41], and we finetune them on our face training set for fair comparisons.”
HiFaceGAN: Face Renovation via Collaborative Suppression and Replenishment (2020, background): “Image restoration consists of a variety of subtasks, such as denoising [1][60], deblurring [25][26] and compression artifact removal [7][39].”
Image Restoration with Mean-Reverting Stochastic Differential Equations (2023, methodology): “For comparison, we report four milestone deblur-ring approaches: DeepDeblur (Nah et al., 2017), Deblur-GAN (Kupyn et al., 2018), DeblurGAN-v2 (Kupyn et al., 2019), DBGAN (Zhang et al., 2020), and MAXIM (Tu et al., 2022).”