Dataset for Paper: Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Vision.
Images: images.tar
dev-labels: llvisionqa_dev.json
test-labels: llvisionqa_test.json
See Github for Usage:
https://github.com/vqassessment/q-bench.
Feel free to cite us.
@article{wu2023qbench,
title={Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Vision},
author={Wu, Haoning and Zhang, Zicheng and Zhang, Erli and Chen, Chaofeng and Liao, Liang and Wang, Annan and… See the full description on the dataset page:
https://huggingface.co/datasets/teowu/LLVisionQA-QBench.
