awesome-deblurring-resources

Real-World Deep Local Motion Deblurring

Intriguing Findings of Frequency Selection for Image Deblurring

Learning Single Image Defocus Deblurring with Misaligned Training Pairs

Dual-Domain Attention for Image Deblurring

Structured Kernel Estimation for Photon-Limited Deconvolution

Blur Interpolation Transformer for Real-World Motion from Blur

Fast Diffusion EM: a diffusion model for blind inverse problems with application to deconvolution

Blind Image Deblurring using FFT-ReLU with Deep Learning Pipeline Integration

Motion Blur Decomposition with Cross-shutter Guidance

Blur-aware Spatio-temporal Sparse Transformer for Video Deblurring

Spike-guided Motion Deblurring with Unknown Modal Spatiotemporal Alignment

A Unified Framework for Microscopy Defocus Deblur with Multi-Pyramid Transformer and Contrastive Learning

Efficient Blind Deblurring under High Noise Levels

Blind Image Deconvolution using Pretrained Generative Priors

Deep Stacked Hierarchical Multi-Patch Network for Image Deblurring

Dynamic Scene Deblurring with Parameter Selective Sharing and Nested Skip Connections

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

The CelebFaces Attributes dataset (CelebA) is a large-scale face attributes dataset comprising 202,599 images of 10,177 celebrities. Each image is 178×218 pixels and annotated with 40 binary labels for facial attributes like hair color, gender, and age.

The Deblur-NeRF dataset focuses on two types of blur: camera motion blur and defocus blur. It includes 5 synthesized scenes for each blur type, created using Blender with multi-view cameras to simulate real data capture. For motion blur, images are rendered from interpolated camera poses, while defocus blur images are generated with depth-of-field effects. Additionally, the dataset features 20 real-world scenes—10 for each blur type—captured with a Canon EOS RP, including both manually blurred images and sharp reference images.

The Dual-Pixel Defocus Deblurring (DPDD) dataset contains 500 carefully captured scenes, comprising 2000 images in total: 500 defocus-blurred images with their 1000 dual-pixel (DP) sub-aperture views and 500 corresponding all-in-focus images, all at full-frame resolution of 6720x4480 pixels.

The GoPro dataset consists of 3,214 pairs of motion-blurred and sharp images, each with a resolution of 1,280×720 pixels, divided into 2,103 training pairs and 1,111 test pairs.

The HIDE (Human-aware Image Deblurring) dataset consists of 8,422 blurred images paired with their corresponding sharp images, focusing on motion deblurring with an emphasis on human subjects, making it ideal for human-centric deblurring tasks.

The RealBlur dataset consists of 4,738 pairs of images from 232 different scenes, captured in both camera raw and JPEG formats. It is divided into two subsets: RealBlur-R with raw images and RealBlur-J with JPEG images, with 3,758 training pairs and 980 test pairs in each subset.

Learning to Deblur using Light Field Generated and Real Defocus Images

Unifying Motion Deblurring and Frame Interpolation with Events

E-CIR: Event-Enhanced Continuous Intensity Recovery

Multi-Scale Memory-Based Video Deblurring

HINet: Half Instance Normalization Network for Image Restoration

Spatio-Temporal Deformable Attention Network for Video Deblurring

Explore Image Deblurring via Encoded Blur Kernel Space

Multi-Stage Progressive Image Restoration

DeFMO: Deblurring and Shape Recovery of Fast Moving Objects

Towards Rolling Shutter Correction and Deblurring in Dynamic Scenes

Digital Gimbal: End-to-end Deep Image Stabilization with Learnable Exposure Times

Rethinking Coarse-to-Fine Approach in Single Image Deblurring

Cascaded Deep Video Deblurring Using Temporal Sharpness Prior

Variational-EM-Based Deep Learning for Noise-Blind Image Deblurring

Deblurring by Realistic Blurring

End-to-end Interpretable Learning of Non-blind Image Deblurring

Efficient Spatio-Temporal Recurrent Neural Network for Video Deblurring

Multi-Temporal Recurrent Neural Networks For Progressive Non-Uniform Single Image Deblurring With Incremental Temporal Training