7 natural image benchmarks
Emerging1papers using it
2026first seen
The '7 natural image benchmarks' dataset contains a collection of diverse natural images used to evaluate the generalization performance of vision-language models across various natural-image domains.