MSIR Bench Review Dataset This repository contains an anonymized review snapshot of MSIR Bench, a benchmark for identity preserving style image retrieval. Dataset Description Each source identity is represented by an anonymous five digit ID. Images are organized by split and identity folder. File names follow either .png , original.png , or legacy original.jpg for original reference images. The dataset is intended for evaluating whether a retrieval model can match images by source identity across visual styles. Splits Split Images Identities : : FLUX.2 Klein 3000 1000 OmniGen2 3000 1000 seenid+unseenstyle 1000 500 train 10000 500 unseenid+seenstyle 9000 3000 unseenid+unseenstyle 9000 3000 Styles Seen styles include: Impressionism , LineArt , InkWash , Gongbi , Ukiyoe , PencilSketch , RoughSketch , Crayon , PopArt , Claymation , Embroidery , Pixar , DigitalPainting , LowPoly , PixelArt , Glitch , Anime , Comic , Caricature . Held out or generated evaluation styles include project defined unseen style candidates such as VintagePhoto , Vaporwave , StreetArt , Bauhaus , ClassicOil , and ChineseCartoon . Privacy and Anonymization Identity folders use anonymous numeric IDs only. JPEG ima…
Runs entirely in your browser via DuckDB-Wasm — this dataset's real data file is loaded once, then queried locally. Nothing is sent to a server.
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy