1.First, confirm that you have sufficient storage space.
df -h
2.Retrieve the content from Github
git clone https://github.com/sunplus-plus1/smzf.git -b master
cd smzf
3.Pulling submodules uses SSH by default. I haven’t configured a key, so I need to change it to HTTPS.
nano .gitmodules
Modify the two URLs:
url = https://github.com/sunplus-plus1/nnthirdparty.git
url = https://github.com/sunplus-plus1/model-zoo.git
4.Next, pull the submodules.
git submodule sync --recursive
git submodule update --init --recursive --remote
5.Modify DNS.
sudo nano /etc/resolv.conf
#Add the following content:
nameserver 8.8.8.8
6.Install CMake.
sudo apt update
sudo apt install cmake
7.Unzip the three files.
cd ./nnthirdparty/libtorch
tar -xzf libtorch.tar.gz
cd ..
cd ffmpeg/
tar -xzf 9.2.tar.gz
cd ..
cd opencv/
tar -xzf 4.2.tar.gz
8.After decompression
cd ..
cd ..
./build.sh --c3v
9.Enter the release directory and use ldd to find missing dependencies.
export LD_LIBRARY_PATH=$(pwd)/lib:$(pwd)/nnthirdparty/jsoncpp/lib:$(pwd)/nnthirdparty/NADKLogger/lib:$(pwd)/nnthirdparty/NADKRoutines/lib:$(pwd)/nnthirdparty/nnbase/lib:$(pwd)/nnthirdparty/libtorch/lib:$LD_LIBRARY_PATH
10.Start using the AI model. Here, we use camera image detection as an example.
./bin/nnModel -m Yolo11sDetectionHybridI8 -v /dev/video42