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")# pip install -U transformers accelerate # 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 Reconsideration of SAM3 Access for Academic Research
Hello SAM3 team,
I recently applied for access to the SAM3 model weights, but my request was not approved. I would like to kindly ask if my application could be reconsidered.
I am a graduate student at Xidian University (Xidian University, Xi'an, China), conducting academic research in computer vision. My research focuses on using video-based foundation models to improve sparse-view 3D reconstruction, including tasks such as foreground/background separation, dynamic object removal, and improving multi-view inputs for 3D Gaussian Splatting reconstruction.
I intend to use SAM3 only for non-commercial academic research. The model will be used locally for experiments, and I will not redistribute the weights or provide any public service based on the model.
If access is granted, I plan to use SAM3 as a research tool and baseline for video segmentation and object masking experiments. Any usage will follow the provided license and citation requirements.
I would greatly appreciate it if you could review my request again. Please let me know if any additional information is required.
Thank you very much for your time and for making SAM3 available to the research community.
Best regards,
Graduate Student, Xidian University
Hugging Face username: UncleKI