search

DAMO-NLP-SG / VCD

[CVPR 2024 Highlight] Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding
Apache License 2.0
196 stars 9 forks source link

can you provide the image subsets you used for evaluation because the whole set of gqa and coco is so large #3

Closed yfzhang114 closed 9 months ago

LengSicong commented 9 months ago

Thanks for your interest.

Unfortunately, we also downloaded the whole dataset and read images from that. You can find all image ids used in COCO and GQA here.