Zevin2023/SEAGULL-100w

Contents Introduction Data Structure Construction Pipeline Acknowledgement Citation Introduction This is a large-scale synthetic ROI-based IQA dataset, SEAGULL-100w, for "SEAGULL: No-reference Image Quality Assessment for Regions of Interest via Vision-Language Instruction Tuning." We create distortion (Dist.) images with six different distortion types, namely blur, sharpness, exposure, contrast, colorfulness and compression. We sample… See the full description on the dataset page: https://huggingface.co/datasets/Zevin2023/SEAGULL-100w.

<iframe src="https://silytics.si/embed/item/d6fe53a3-a5cb-4a9c-9f45-b2fbae8b2c9f" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.