Real-World Blur Dataset for Learning and Benchmarking Deblurring Algorithms

By Jaesung Rim, H. Chwa, Sunghyun Cho. European Conference on Computer Vision, 2020

492 citations recorded by the source corpus.

Numerous learning-based approaches to single image deblurring for camera and object motion blurs have recently been proposed. To generalize such approaches to real-world blurs, large datasets of real blurred images and their ground truth sharp images are essential. However, there are still no such datasets, thus all the existing approaches resort to synthetic ones, which leads to the failure of deblurring real-world images. In this work, we present a large-scale dataset of real-world blurred images and their corresponding sharp images captured in low-light environments for learning and benchmarking single image deblurring methods. To collect our dataset, we build an image acquisition system to simultaneously capture a geometrically aligned pair of blurred and sharp images, and develop a post-processing method to further align images geometrically and photometrically. We analyze the effect of our post-processing step, and the performance of existing learning-based deblurring methods. Our analysis shows that our dataset significantly improves deblurring quality for real-world low-light images.

View Paper: https://www.semanticscholar.org/paper/2d6c14023087b5d5bd90a88da13e0fa765418d84

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Deep Generalized Unfolding Networks for Image Restoration (2022, methodology): “For image deblurring, similar to [36, 55, 63, 79, 80], we train our model with 2, 103 image pairs from GoPro [47] dataset and evaluate each method on the test sets from GoPro and HIDE [58], which contain 1, 111 and 2, 025 samples, respectively.”
Deep Image Deblurring: A Survey (2022, methodology): “For a better understanding of the role of different training datasets, we train the SRN model Tao et al. (2018) on three different public datasets (RealBlur-J Rim et al. 2020, GoPro Nah et al. 2017, and BSD-B Martin et al. 2001; Rim et al. 2020) separately.”
Rethinking Coarse-to-Fine Approach in Single Image Deblurring (2021, methodology): “We used the GoPro [20] and RealBlur [25] training datasets for training our models which consist of 2,103 and 3,758 pairs of blurred and sharp images.”
Multi-Stage Progressive Image Restoration (2021, methodology): “We evaluate our MPRNet on the real-world images of a recent RealBlur [64] dataset under two experimental settings: 1).”
Exploring CLIP for Assessing the Look and Feel of Images (2022, methodology): “We then extend our scope to finegrained attributes using synthetic data and common restoration benchmarks (Wei et al. 2018; Bychkovsky et al. 2011; Xu et al. 2018; Rim et al. 2020).”
Stripformer: Strip Transformer for Fast Image Deblurring (2022, methodology): “To address real-world blurs, we evaluate our method on the RealBlur dataset [35], which has 3, 758 blurred and sharp pairs for training and 980 pairs for testing.”
Efficient Frequency Domain-based Transformers for High-Quality Image Deblurring (2022, methodology): “We evaluate our method on commonly used image deblurring datasets including the GoPro dataset [13], the HIDE dataset [17], and the RealBlur dataset [15].”
MAXIM: Multi-Axis MLP for Image Processing (2022, methodology): “Deblurring GoPro [60] 2103 1111 GoPro HIDE [79] 0 2025 HIDE RealBlur-J [77] 3758 980 RealBlur-J RealBlur-R [77] 3758 980 RealBlur-R REDS [61] 24000 300 REDS”
Uformer: A General U-Shaped Transformer for Image Restoration (2021, methodology): “Then we apply our GoPro trained model directly on the HIDE dataset [49] and the RealBlur dataset [45] to evaluate the generalization on real scenes.”
Deblurring via Stochastic Refinement (2021, methodology): “This was done to make qualitative evaluation in a more controlled environment, since the low-quality ground truth images in existing paired datasets [50, 56] make qualitative assessment difﬁcult.”
Restormer: Efficient Transformer for High-Resolution Image Restoration (2021, methodology): “Our Restormer is trained only on the GoPro dataset [56] and directly applied to the HIDE [67] and RealBlur [65] benchmark datasets.”
Efficient and Explicit Modelling of Image Hierarchies for Image Restoration (2023, methodology): “3 shows the experimental results for single image motion deblur-ring on synthetic datasets (GoPro [51], HIDE [59]) and real dataset (RealBlur-R [57]), respectively.”