Instructions to use facebook/sam3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use facebook/sam3 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("mask-generation", model="facebook/sam3")# Load model directly from transformers import AutoProcessor, AutoModel processor = AutoProcessor.from_pretrained("facebook/sam3") model = AutoModel.from_pretrained("facebook/sam3", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Request for Manual Reset of Rejected SAM 3 Access
Hello Meta SAM 3 Team,
My access request for the facebook/sam3 repository was recently rejected. I am writing to kindly request a manual review or a reset of my application status so I can reapply.
I am a graduate student at University of Arizona conducting non-commercial academic research in the field of computer vision. My current project focuses on 3D human pose estimation, where I aim to evaluate the performance of SAM 3 on human body segmentation and keypoint detection tasks. Specifically, I plan to use SAM 3 to generate high-quality segmentation masks of human bodies from video sequences, which will serve as input for downstream pose estimation models.
My intended use of SAM 3 is strictly for academic research and experimentation purposes only. The model will be run locally on our lab's research workstations, and I will not redistribute the model weights or use it for any commercial applications.
If my previous request was incomplete or unclear, I would appreciate the opportunity to correct it. I confirm that I will fully comply with the SAM License, applicable laws, the Meta Privacy Policy, and Hugging Face's terms of use.
My Hugging Face username is: Lidonglong1
Thank you for your time and for sharing this remarkable work with the community.