Skip to content

Latest commit

 

History

History
39 lines (30 loc) · 1.44 KB

File metadata and controls

39 lines (30 loc) · 1.44 KB

Face detection demo

OpenVINO™ Model Server fork of MediaPipe repository allowing users to take advantage of OpenVINO™ backend in mediapipe framework.

Original demo documentation

Building docker container with dependencies

git clone https://github.com/openvinotoolkit/mediapipe.git
cd mediapipe
make docker_build

Start mediapipe_ovms container

docker run -it mediapipe_ovms:latest bash

Running demo inside mediapipe_ovms container

Prepare models and build application

python setup_ovms.py --get_models
python setup_ovms.py --convert_pose --force
bazel build -c opt --define MEDIAPIPE_DISABLE_GPU=1 mediapipe/examples/desktop/face_detection:face_detection_cpu

Download the input video or prepare your own input as 'video.mp4'

wget -O video.mp4 "https://www.pexels.com/download/video/3044127/?fps=24.0&h=1080&w=1920"

Run the demo

Execute command with the calculator_graph_config_file, input_video_path and output_video_path:

bazel-bin/mediapipe/examples/desktop/face_detection/face_detection_cpu --calculator_graph_config_file /mediapipe/mediapipe/graphs/face_detection/face_detection_desktop_live.pbtxt --input_video_path=/mediapipe/video.mp4 --output_video_path=/mediapipe/output_face_detection_ovms.mp4

Now you can review the output_face_detection_ovms.mp4 output with detections.