Awesome
Fast Segment Anything
[📕Paper
] [🤗HuggingFace Demo
] [Colab demo
] [Replicate demo & API
] [Model Zoo
] [BibTeX
]
The Fast Segment Anything Model(FastSAM) is a CNN Segment Anything Model trained by only 2% of the SA-1B dataset published by SAM authors. The FastSAM achieve a comparable performance with the SAM method at 50× higher run-time speed.
🍇 Refer from
https://github.com/CASIA-IVA-Lab/FastSAM
[[Original
]((https://github.com/CASIA-IVA-Lab/FastSAM)]
Export ONNX
# You can re-config pt2onnx.sh
bash pt2onnx.sh
or
python pt2onnx.py
INFER ONNX
python infer_onnx.py
<img src="images/cat.jpg" width="300" height="600" />
<img src="outputs/obj.png" width="300" height="600"/>
TensorRT (You can deploy on Triton Server by tensort plan)
bash onnx2trt.sh
python inference_trt.py
<img src="images/cong-torii-noi-tieng-nhat-ban.jpg" width="500" />
<img src="outputs/obj_trt.png" width="500" />