Super Fire Mini GPT-4's visual ability skyrocketed, GitHub 20000 stars, produced by a Chinese team
Bai Jiao originates from Aofei TempleQuantum bit | official account QbitAIGPT-4V for object detection? Netizens tested: Not ready yet.Although the detected categories are fine, most bounding boxes are misplaced
Bai Jiao originates from Aofei Temple
Quantum bit | official account QbitAI
GPT-4V for object detection? Netizens tested: Not ready yet.
Although the detected categories are fine, most bounding boxes are misplaced.
It's okay, someone will take action!
MiniGPT-v2.
And it's just a simple command:[grounding] describe this image in detail.
.
[identify] .
Of course, you can also ask without adding anything~
MiniGPT-v2MiniGPT-4KAUSTMeta.
MiniGPT-4GItHub22000+.
This upgrade has already been used by netizens~
Universal interface for multiple visual tasks
..
.
MiniGPT-v2.
ViT.ViT LLaMA-2.
.
.
MiniGPT-v2 .
.
Interested friends can click on the demo link below to experience:
https://minigpt-v2.github.io/
https://huggingface.co/spaces/Vision-CAIR/MiniGPT-v2
Paper link: https://arxiv.o Rg/abs/2310.09478
GitHub link: https://github.com/Vision-CAIR/MiniGPT-4
Reference link: https://twitter.com/leoyerrrr
Tag: Super Fire Mini GPT-4 visual ability skyrocketed GitHub 20000
Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.
