Benchmark Script for FastSAM and ORB

Indice

Requisiti

Piattaforme supportate

Setup ambiente

Linux / macOS

python3 -m venv venv
source venv/bin/activate

Windows (PowerShell)

python -m venv venv
venv\Scripts\Activate.ps1

Installazione PyTorch

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124

pip install torch torchvision torchaudio

Installazione dipendenze

pip install -r requirements.txt

pip install opencv-python-headless

Struttura progetto

project/
├── benchmark_fastsam_orb.py     # Script principale (no-args)
├── requirements.txt
├── README.md
├── FastSAM-x.pt                  # (opzionale: lo script può scaricarli)
├── FastSAM-s.pt                  # (opzionale: lo script può scaricarli)
└── images/
    ├── image1.jpg
    ├── image2.jpg
    └── image3.jpg

Uso

python benchmark_fastsam_orb.py

Uso su Jetson (Nano/Orin/Xavier)

1) Ambiente & dipendenze

python3 -m venv venv
source venv/bin/activate
pip install --upgrade pip wheel
pip install opencv-python-headless numpy pandas pillow tqdm

2) PyTorch per Jetson

wget https://nvidia.box.com/shared/static/3nm3gxxwy8f8q4nwd6kgtl4p3yr3zd8i.whl -O torch-1.10.0-cp38-cp38-linux_aarch64.whl
pip install torch-1.10.0-cp38-cp38-linux_aarch64.whl
pip install torchvision==0.11.1

wget https://nvidia.box.com/shared/static/bz5lxprp6k25lhk1pnc6u9h7v4rfv74i.whl -O torch-2.1.0-cp38-cp38-linux_aarch64.whl
pip install torch-2.1.0-cp38-cp38-linux_aarch64.whl
pip install torchvision==0.16.0

Ottimizzazioni & consigli

Troubleshooting