yolo-object-detection-cli

Usage

Docker

Pull dependencies

git lfs clone https://github.com/aslafy-z/yolo-object-detection-samples ./samples
docker pull ghcr.io/aslafy-z/yolo-object-detection-cli:main

Run with CPU support:

docker run -it --rm \
  -v $PWD/samples/data:/samples \
  -p 8000:8000 \
  ghcr.io/aslafy-z/yolo-object-detection-cli:main \
  --source=/samples/shop.mp4

Run with GPU support:

docker run -it --rm \
  -v $PWD/samples/data:/samples \
  -p 8000:8000 \
  --ipc=host --gpus=all --runtime=nvidia \
  ghcr.io/aslafy-z/yolo-object-detection-cli:main \
  --source=/samples/shop.mp4 \
  --device=cuda:0

System

sudo apt update && sudo apt install -y ffmpeg libsm6 libxext6

git clone https://github.com/aslafy-z/yolo-object-detection-cli && cd ./yolo-object-detection-cli
git lfs clone https://github.com/aslafy-z/yolo-object-detection-samples ./samples

python -m virtualenv venv
source ./venv/bin/activate
pip install -r requirements.txt

python main.py --source=samples/data/shop.mp4
python main.py --source=rtsp://user:pass@10.0.0.5/Src/MediaInput/stream_1

Development

Lock dependencies

uv lock
uv pip compile pyproject.toml -o requirements.txt

Note: Docker image uses a subset of the dependencies, see in requirements.docker.txt.

Run

git lfs clone https://github.com/aslafy-z/yolo-object-detection-samples ./samples
uv run main.py --source=samples/data/shop.mp4