mlsanityData to 14 Oct 2026

Attentive Fine-Grained Structured Sparsity for Image Restoration

arXiv 2204.12266 ↗First posted 26 Apr 2022cs.CVcited by 12

Not on the map. Why?

Abstract (arXiv)

Image restoration tasks have witnessed great performance improvement in recent years by developing large deep models. Despite the outstanding performance, the heavy computation demanded by the deep models has restricted the application of image restoration. To lift the restriction, it is required to reduce the size of the networks while maintaining accuracy. Recently, N:M structured pruning has appeared as one of the effective and practical pruning approaches for making the model efficient with the accuracy constraint. However, it fails to account for different computational complexities and performance requirements for different layers of an image restoration network. To further optimize the trade-off between the efficiency and the restoration accuracy, we propose a novel pruning method that determines the pruning ratio for N:M structured sparsity at each layer. Extensive experimental results on super-resolution and deblurring tasks demonstrate the efficacy of our method which outperforms previous pruning methods significantly. PyTorch implementation for the proposed methods is available at https://github.com/JungHunOh/SLS_CVPR2022.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Builds on 51 of its 66 references that are in the map. First posted on arXiv 26 Apr 2022.

Builds on

51 of 66 references.

References · 51

  1. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  2. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,329
  3. 2018The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networkscited by 2,262
  4. 2015Learning both Weights and Connections for Efficient Neural Networkscited by 2,031
  5. 2017Enhanced Deep Residual Networks for Single Image Super-Resolutioncited by 1,552
  6. 2016Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learningcited by 1,512
  7. 2018Image Super-Resolution Using Very Deep Residual Channel Attention Networkscited by 1,264
  8. 2018ESRGAN: Enhanced Super-Resolution Generative Adversarial Networkscited by 1,254
  9. 2014Image Super-Resolution Using Deep Convolutional Networkscited by 1,224
  10. 2016Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Networkcited by 1,204
  11. 2016Pruning Filters for Efficient ConvNetscited by 1,143
  12. 2015Accurate Image Super-Resolution Using Very Deep Convolutional Networkscited by 903
38 more
  1. 2021Multi-Stage Progressive Image Restorationcited by 816
  2. 2018Residual Dense Network for Image Super-Resolutioncited by 799
  3. 2016Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurringcited by 727
  4. 2017Channel Pruning for Accelerating Very Deep Neural Networkscited by 702
  5. 2017Learning Efficient Convolutional Networks through Network Slimmingcited by 676
  6. 2016Learning Structured Sparsity in Deep Neural Networkscited by 541
  7. 2019Importance Estimation for Neural Network Pruningcited by 541
  8. 2017ThiNet: A Filter Level Pruning Method for Deep Neural Network Compressioncited by 442
  9. 2018Filter Pruning via Geometric Median for Deep Convolutional Neural Networks Accelerationcited by 367
  10. 2019DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Bettercited by 362
  11. 2018Scale-recurrent Network for Deep Image Deblurringcited by 333
  12. 2020HRank: Filter Pruning using High-Rank Feature Mapcited by 292
  13. 2020Single Image Super-Resolution via a Holistic Attention Networkcited by 265
  14. 2018Fast, Accurate, and Lightweight Super-Resolution with Cascading Residual Networkcited by 264
  15. 2019What's Hidden in a Randomly Weighted Neural Network?cited by 237
  16. 2019MetaPruning: Meta Learning for Automatic Neural Network Channel Pruningcited by 227
  17. 2018Deep Back-Projection Networks For Super-Resolutioncited by 225
  18. 2021Accelerating Sparse Deep Neural Networkscited by 223
  19. 2019Deep Stacked Hierarchical Multi-patch Network for Image Deblurringcited by 206
  20. 2021Learning N:M Fine-grained Structured Sparse Neural Networks From Scratchcited by 179
  21. 2019Towards Optimal Structured CNN Pruning via Generative Adversarial Learningcited by 146
  22. 2020Residual Feature Distillation Network for Lightweight Image Super-Resolutioncited by 122
  23. 2020Spatially-Attentive Patch-Hierarchical Network for Adaptive Motion Deblurringcited by 119
  24. 2019Gate Decorator: Global Filter Pruning Method for Accelerating Deep Convolutional Neural Networkscited by 118
  25. 2018Rethinking the Smaller-Norm-Less-Informative Assumption in Channel Pruning of Convolution Layerscited by 112
  26. 2019Multi-Temporal Recurrent Neural Networks For Progressive Non-Uniform Single Image Deblurring With Incremental Temporal Trainingcited by 112
  27. 2021Accelerated Sparse Neural Training: A Provable and Efficient Method to Find N:M Transposable Maskscited by 103
  28. 2021ClassSR: A General Framework to Accelerate Super-Resolution Networks by Data Characteristiccited by 81
  29. 2020EagleEye: Fast Sub-net Evaluation for Efficient Neural Network Pruningcited by 80
  30. 2020Operation-Aware Soft Channel Pruning using Differentiable Maskscited by 74
  31. 2020DMCP: Differentiable Markov Channel Pruning for Neural Networkscited by 69
  32. 2019Fast, Accurate and Lightweight Super-Resolution with Neural Architecture Searchcited by 66
  33. 2020DSA: More Efficient Budgeted Pruning via Differentiable Sparsity Allocationcited by 47
  34. 2019Efficient Residual Dense Block Search for Image Super-Resolutioncited by 37
  35. 2018Balanced Sparsity for Efficient DNN Inference on GPUcited by 27
  36. 2021SRWarp: Generalized Image Super-Resolution under Arbitrary Transformationcited by 27
  37. 2021Clean Images are Hard to Reblur: Exploiting the Ill-Posed Inverse Task for Dynamic Scene Deblurringcited by 17
  38. 2019Fine-Grained Neural Architecture Searchcited by 6

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year

Citing papers per year: 2022: 3; 2023: 5; 2024: 3; 2025: 1.

12 papers.

  1. 2023OTOV2: Automatic, Generic, User-Friendlycited by 32
  2. 2022CADyQ: Content-Aware Dynamic Quantization for Image Super-Resolutioncited by 27
  3. 2023Why is the State of Neural Network Pruning so Confusing? On the Fairness, Comparison Setup, and Trainability in Network Pruningcited by 16
  4. 2022Structured Sparsity Learning for Efficient Video Super-Resolutioncited by 14
  5. 2024AdaBM: On-the-Fly Adaptive Bit Mapping for Image Super-Resolutioncited by 9
  6. 2024Dynamic Contrastive Knowledge Distillation for Efficient Image Restorationcited by 6
  7. 2023Overcoming Distribution Mismatch in Quantizing Image Super-Resolution Networkscited by 5
  8. 2023Iterative Soft Shrinkage Learning for Efficient Image Super-Resolutioncited by 2
  9. 2023Distribution-Flexible Subset Quantization for Post-Quantizing Super-Resolution Networkscited by 2
  10. 2024Cross-Block Fine-Grained Semantic Cascade for Skeleton-Based Sports Action Recognitioncited by 1
  11. 2025Deep Learning Empowered Super-Resolution: A Comprehensive Survey and Future Prospectscited by 1
  12. 2022ClassPruning: Speed Up Image Restoration Networks by Dynamic N:M Pruningcited by 0