Conda
啟動 conda 環境: conda activate aiot
Central broker
打開 docker console ，找到 rabbitmq 點選 start
回到程式資料夾底下，啟動 central broker
Config
刪除或新增"APP"的UI介面，上傳 config
啟動：
python main.py
上傳：
POST -> try it out -> 選擇檔案 -> execute
查看 Responses -> Code，如果有成功會寫: Build App successfully
Server
到 application 的資料夾底下，啟動 server
python server.py
Client
到 application 的資料夾底下，用 client.py 送資料
python client.py --audio <filename>.wav
Build container from a dockerfile
docker build -t <name> .
docker run --rm <name>
git clone https://github.com/openvinotoolkit/open_model_zoo.git
cd open_model_zoo\demos
cd multi_camera_multi_target_tracking_demo\python
conda create --name openvino python=3.7
conda activate openvino
pip install openvino
cd ..\..
pip install -r requirements.txt
cd multi_camera_multi_target_tracking_demo\python
pip install openvino-dev
pip install --upgrade pip
pip install .
omz_downloader --list models.lst
omz_converter --list models.lst
python multi_camera_multi_target_tracking_demo.py -i "\cam1.mp4" "\cam4.mp4" --m_detector "\open_model_zoo\demos\multi_camera_multi_target_tracking_demo\python\intel\person-detection-retail-0013\FP16\person-detection-retail-0013.xml" --m_reid "\open_model_zoo\demos\multi_camera_multi_target_tracking_demo\python\intel\person-reidentification-retail-0277\FP16\person-reidentification-retail-0277.xml" --config configs\person.py --output_video outout.avi
有執行結果，但完全不準 = W =
ref:
https://github.com/openvinotoolkit/open_model_zoo/tree/master/demos/multi_camera_multi_target_tracking_demo/python
git clone https://github.com/KaiyangZhou/deep-person-reid.git
# method 1
conda create --name cmuOD python=3.7
conda activate cmuOD
pip install -r requirements.txt
conda install tensorflow-gpu==1.15
python obj_detect_tracking.py --model_path obj_v3_model --version 3 --video_dir v1-val_testvideos --video_lst_file v1-val_testvideos.lst --frame_gap 1 --get_tracking --tracking_dir test_track_out # can't successfully execute yet
(ref: 最簡單的python Tensorflow-gpu安裝方法)
# method 2
conda create --name cmuOD python=3.7
conda activate cmuOD
conda install -c conda-forge cudatoolkit=10.0 cudnn=7.6.5
pip install -r requirements.txt
pip install tensorflow-gpu==1.15
# test tensorflow with GPU
from tensorflow.python.client import device_lib
print(device_lib.list_local_devices())
# requirements.txt
numpy==1.19
scipy
scikit-learn
opencv-python
matplotlib
pycocotools
tqdm
protobuf==3.20.*
psutil
pyyaml
link: https://github.com/daedalus-tech/mc-mot
conda create -n mc-mot python=3.10
conda activate mc-mot
pip install torch==1.12.0+cu113 torchvision==0.13.0+cu113 torchaudio==0.12.0 --extra-index-url https://download.pytorch.org/whl/cu113
pip install opencv-python
pip install pandas
pip install psutil
pip install pyyaml
pip install tqdm
pip install ultralytics
pip install filterpy
python calibrate.py --video1 .\cam1.mp4 --video2 .\cam4.mp4 --homography-pth .\homography
python main.py --video1 .\cam1.mp4 --video2 .\cam4.mp4 --homography .\homography.npy
git clone https://github.com/stefanopini/simple-HRNet.git
cd simple-HRNet
conda env remove -n hrnet
conda create -n hrnet python=3.9
conda activate hrnet
pip install -r requirements.txt
pip install torch==1.12.0+cu113 torchvision==0.13.0+cu113 torchaudio==0.12.0 --extra-index-url https://download.pytorch.org/whl/cu113
# demo
python scripts/live-demo.py --camera_id 0 --single_person --device cuda
python scripts/live-demo.py --filename video.mp4 --single_person --disable_vidgear --save_video
1. 先檢查麥克風熱鍵(F9)是否開啟
2. 如開啟後仍無聲音，檢查麥克風設定。如：Google Meet 中，設定 > 音訊 > 麥克風 要選擇 Microphone Array 而不是預設的 Microphone；Skype 同理也可解決
不知為何預設的音訊來源不能用，記得不管是哪個軟體或程式，把麥克風音訊的來源裝置設定一下即可。
# 用特定路徑底下的 dockerfile build image
docker build -t welcome-to-docker .
# 多個 container 組在一起啟動
docker compose up -d
# 查看有哪些 images
docker images
