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 reconsideration of rejected SAM3 access
#325
by geh-2287 - opened
Hello Meta SAM3 team,
My request to access facebook/sam3 on Hugging Face was rejected.
Could you please reconsider my request or reset my rejected access request so that I can submit corrected information?
My Hugging Face username:
- geh-2287
Affiliation:
- Independent developer
Intended use:
- I am working on a non-commercial secondary development project for a Unitree G1 robot.
- The goal is to build a local robotics perception prototype using an external Intel RealSense RGB-D camera.
- I would like to use SAM3 for text-prompted object segmentation.
- The segmentation mask will be combined with RealSense depth to estimate the target object's 3D position in the camera frame.
- The initial prototype is read-only perception only and does not involve autonomous robot motion or public deployment of the model weights.
Use case details:
- Input: RGB images from an external RealSense camera.
- Task: text-prompted object segmentation, such as locating objects like a desk or other scene objects.
- Output: object masks and camera-frame 3D position estimates using depth.
- Environment: local development and testing.
License and compliance:
- This is for non-commercial robotics prototyping.
- I will comply with the model license and usage restrictions.
- I will not redistribute the model weights.
- I will not use the model for prohibited or harmful applications.
If my original request was missing information, I would appreciate the opportunity to provide more complete details.
Thank you.