重思超分辨率训练数据，发现高质量多样数据能显著提升模型性能
Rethinking Image Super-Resolution from Training Data Perspectives
- 构建自动化评估流水线，按质量和多样性筛选图像数据
- 低压缩伪影、多物体场景、来自ImageNet/PASS的大规模数据提升效果
- 为未来超分数据集构建提供实用指导，适合数据工程师和模型开发者
本文从多样性和质量角度重新审视图像超分辨率（SR）的训练数据。当前多数新方法在DIV2K和DF2K等常见数据集上开发与评测，但训练数据对模型性能的影响未被充分研究。为此，我们提出一种自动化图像评估流水线，对现有高分辨率数据集以及大规模图像数据集如ImageNet和PASS进行分层比较。结果表明：(i) 压缩伪影少的数据更优，(ii) 图像内物体种类多（即多样性高）的数据表现更好，(iii) 来自ImageNet或PASS的大量图像有助于提升性能。该简单而有效的数据筛选流程可为未来超分辨率数据集建设提供参考，推动整体模型进步。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。