Jetson Nano Inference experiments
blindly following AI on the Jetson Nano
- This is very likely cloned from Dusty README.md file ...
-
sudo apt-get install git cmake libpython3-dev python3-numpy
-
cd ~/Downloads
-
git clone --recursive https://github.com/dusty-nv/jetson-inference
-
cd jetson-inference
-
mkdir build
-
cd build
-
cmake ../
-
When presented with selection dialog boxes,
-
Select ALL pre-trained Image Recognition models (2.2GB)
-
Select ALL detection models (395MB)
-
Take SEMANTIC Segmentation 24.25,27,29,31,33 (not legacy models)
-
Take ALL Image-Processing -ALL models
-
Also -- select PyTorch 3.X version for installation
-
make -j$(nproc)
-
sudo make install
-
sudo ldconfig
-
sudo apt-get install v4l-utils
- test with loading Python3 and 'import torch' 'import torchvision' and they should load without error
- quit()
create quick test program with
-
n_net = jetson_inference.detectNet('ssd-mobilenet-v2', threshold=0.5)
-
n_net = jetson_inference.imageNet('googlenet')