Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: other
|
| 3 |
+
license_name: sam-license
|
| 4 |
+
license_link: LICENSE
|
| 5 |
+
tags:
|
| 6 |
+
- segmentation
|
| 7 |
+
- image-segmentation
|
| 8 |
+
- video-segmentation
|
| 9 |
+
- sam3
|
| 10 |
+
- meta
|
| 11 |
+
- pytorch
|
| 12 |
+
- safetensors
|
| 13 |
+
pipeline_tag: image-segmentation
|
| 14 |
+
base_model: facebook/sam3
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
# SAM3 — Segment Anything Model 3
|
| 18 |
+
|
| 19 |
+
This is an **ungated mirror** of Meta's [SAM3](https://github.com/facebookresearch/sam3) model weights, redistributed under the [SAM License](LICENSE) for easier access.
|
| 20 |
+
|
| 21 |
+
## What is SAM3?
|
| 22 |
+
|
| 23 |
+
SAM3 (Segment Anything Model 3) is Meta AI's state-of-the-art foundation model for **promptable segmentation** in images and videos. It supports:
|
| 24 |
+
|
| 25 |
+
- **Text prompts** — segment objects by describing them (e.g. *"the dog"*)
|
| 26 |
+
- **Bounding box prompts** — precise region targeting
|
| 27 |
+
- **Point prompts** — click-to-segment
|
| 28 |
+
- **Video tracking** — track and segment objects across video frames
|
| 29 |
+
- **Automatic mask generation** — segment all objects in a scene
|
| 30 |
+
|
| 31 |
+
## Files
|
| 32 |
+
|
| 33 |
+
| File | Description |
|
| 34 |
+
|---|---|
|
| 35 |
+
| `sam3.safetensors` | Model weights (safetensors format) |
|
| 36 |
+
| `LICENSE` | SAM License (required for redistribution) |
|
| 37 |
+
|
| 38 |
+
## Usage
|
| 39 |
+
|
| 40 |
+
```python
|
| 41 |
+
# With the official sam3 package
|
| 42 |
+
pip install sam3
|
| 43 |
+
|
| 44 |
+
from sam3.model_builder import build_sam3_image_model
|
| 45 |
+
model = build_sam3_image_model(checkpoint_path="sam3.safetensors")
|
| 46 |
+
```
|
| 47 |
+
|
| 48 |
+
## License
|
| 49 |
+
|
| 50 |
+
This model is distributed under the **SAM License** — see the [LICENSE](LICENSE) file. Key points:
|
| 51 |
+
|
| 52 |
+
- ✅ Commercial use permitted
|
| 53 |
+
- ✅ Redistribution permitted (with license included)
|
| 54 |
+
- ✅ Derivative works permitted
|
| 55 |
+
- ❌ No military/warfare, nuclear, or espionage use
|
| 56 |
+
- ❌ No reverse engineering
|
| 57 |
+
|
| 58 |
+
## Credits
|
| 59 |
+
|
| 60 |
+
- **Original model by**: [Meta AI (FAIR)](https://github.com/facebookresearch/sam3)
|
| 61 |
+
- **Original HuggingFace repo**: [facebook/sam3](https://huggingface.co/facebook/sam3)
|
| 62 |
+
- **Paper**: *Segment Anything Model 3* (submitted to ICLR 2026)
|
| 63 |
+
- **Redistributed by**: [Æmotion Studio](https://huggingface.co/AEmotionStudio) for use with [ComfyUI-FFMPEGA](https://github.com/AEmotionStudio/ComfyUI-FFMPEGA)
|