Ferret: Refer and Ground Anything Anywhere at Any Granularity
A multimodal language model that connects conversational language with points, boxes, and free-form image regions.
BibTeX citation
@inproceedings{ferret2024,
title = {Ferret: Refer and Ground Anything Anywhere at Any Granularity},
author = {Haoxuan You and Haotian Zhang and Zhe Gan and Xianzhi Du and Bowen Zhang and Zirui Wang and Liangliang Cao and Shih-Fu Chang and Yinfei Yang},
booktitle = {ICLR},
year = {2024},
url = {https://arxiv.org/abs/2310.07704}
}