mlsanityData to 7 Oct 2026

Rethinking Image Super-Resolution from Training Data Perspectives

arXiv 2409.00768 ↗First posted 1 Sep 2024cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

In this work, we investigate the understudied effect of the training data used for image super-resolution (SR). Most commonly, novel SR methods are developed and benchmarked on common training datasets such as DIV2K and DF2K. However, we investigate and rethink the training data from the perspectives of diversity and quality, {thereby addressing the question of ``How important is SR training for SR models?''}. To this end, we propose an automated image evaluation pipeline. With this, we stratify existing high-resolution image datasets and larger-scale image datasets such as ImageNet and PASS to compare their performances. We find that datasets with (i) low compression artifacts, (ii) high within-image diversity as judged by the number of different objects, and (iii) a large number of images from ImageNet or PASS all positively affect SR performance. We hope that the proposed simple-yet-effective dataset curation pipeline will inform the construction of SR datasets in the future and yield overall better models.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 23 of its 41 references that are in the map. First posted on arXiv 1 Sep 2024.

Builds on

23 of 41 references.

References · 23

  1. 2023Segment Anythingcited by 8,158
  2. 2016Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Networkcited by 1,898
  3. 2021SwinIR: Image Restoration Using Swin Transformercited by 1,869
  4. 2017Enhanced Deep Residual Networks for Single Image Super-Resolutioncited by 1,543
  5. 2018Image Super-Resolution Using Very Deep Residual Channel Attention Networkscited by 1,256
  6. 2018ESRGAN: Enhanced Super-Resolution Generative Adversarial Networkscited by 1,250
  7. 2016Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Networkcited by 1,199
  8. 2015Accurate Image Super-Resolution Using Very Deep Convolutional Networkscited by 902
  9. 2020Pre-Trained Image Processing Transformercited by 891
  10. 2018Residual Dense Network for Image Super-Resolutioncited by 796
  11. 2022Detecting Twenty-thousand Classes using Image-level Supervisioncited by 613
  12. 2022Activating More Pixels in Image Super-Resolution Transformercited by 433
11 more
  1. 2016Accelerating the Super-Resolution Convolutional Neural Networkcited by 393
  2. 2015Sketch-based Manga Retrieval using Manga109 Datasetcited by 326
  3. 2020Single Image Super-Resolution via a Holistic Attention Networkcited by 265
  4. 2019Residual Non-local Attention Networks for Image Restorationcited by 236
  5. 2022Efficient Long-Range Attention Network for Image Super-resolutioncited by 151
  6. 2020Cross-Scale Internal Graph Neural Network for Image Super-Resolutioncited by 94
  7. 2022SwinFIR: Revisiting the SwinIR with Fast Fourier Convolution and Improved Training for Image Super-Resolutioncited by 67
  8. 2021On Efficient Transformer-Based Image Pre-training for Low-Level Visioncited by 64
  9. 2021PASS: An ImageNet replacement for self-supervised pretraining without humanscited by 51
  10. 2022Revisiting RCAN: Improved Training for Image Super-Resolutioncited by 32
  11. 2023HQ-50K: A Large-scale, High-quality Dataset for Image Restorationcited by 15

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 3.

6 papers.

  1. 2025Emulating Self-attention with Convolution for Efficient Image Super-Resolutioncited by 5
  2. 2025LR0.FM: Low-Res Benchmark and Improving Robustness for Zero-Shot Classification in Foundation Modelscited by 3
  3. 2025Instance Data Condensation for Image Super-Resolutioncited by 1
  4. 2026Rank-Factorized Implicit Neural Bias: Scaling Super-Resolution Transformer with FlashAttentioncited by 1
  5. 20264KLSDB: A Large-Scale Dataset for 4K Image Restoration and Generationcited by 0
  6. 2026CAT-Flow: Curvature-Adaptive sTeps for Flow Matchingcited by 0