一、官方資料與說明

網址：https://github.com/dusty-nv/jetson-inference/blob/master/docs/building-repo-2.md#downloading-models

Building the Project from Source

Provided with the repo is a library of TensorRT-accelerated deep learning networks for image recognition, object detection with localization (i.e. bounding boxes), and semantic segmentation. This inferencing library (libjetson-inference) is intended to be run on the Jetson, and includes support for both C++ and Python. Various pre-trained DNN models are automatically downloaded to get you up and running quickly.

The latest source code or Docker container can be used onboard your Jetson once your device has been flashed with JetPack or setup with the pre-populated SD card image. You can run the pre-built Docker container without needing to compile the project or install PyTorch yourself.

我看照這一句之後，我覺得....可能我自己是可以不用去compiler這個Project啦

所以....我直接下載模型就好，而且Docker container已經有Pytorch了，我就....嗯

目前我是已經離開Docker container的狀態下去執行安裝的喔!

執行Docker檢視下載的飛機辨識模型吧

切換到 /jetson-inference/build/aarch64/bin 資料夾

再切換到 networks 資料夾

下 ls 指令

測試看看吧，先上網下載兩張照片，儲存在images資料夾

我們下載的模型是detectnet的，所以就不是使用 imagenet.py，要改用 detectnet-console.py

# Python

$  ./detectnet-console.py  images/a01.jpeg images/test/output_a01.jpg

$  ./detectnet-console.py  images/a02.jpg images/test/output_a02.jpg