Hybrid AI model for Efficient Edge AI for Security.
111
Requirements: The following requirements will need to be met to run the docker container.
Pulling the repository:
docker pull telepathboy/eeai_hybrid:hybrid_model
Build the docker container image:
docker build -t <image_name> .
Run Docker Contianer:
docker run --gpus all --shm-size 8G -it <image_name>
To run DeepSORT example cd into deep_sort directory
cd deep_sort
run deep_sort code:
python3 run_full_deep_sort.py --video=<path_to_your_video_file> --model=resources/networks/mars-small128.pb --min_confidence=0.3 --nn_budget=100 --display=False --save_video=True
To run DeepSORT on a folder of video files. First, populate the Videos folder
python3 run_full_deep_sort.py --video_folder=Videos
To run POCO example: cd into POCO directory
cd POCO
Run POCO code:
python3 demo.py --mode video --vid_file demo_data/ShootingInOTR.mp4 --cfg configs/demo_poco_cliff.yaml --ckpt data/poco_cliff.pt --output_folder out
To run POCO on a folder of video files. First, populate the Videos folder.
python3 run_poco.py --mode video --video_folder Videos --cfg configs/demo_poco_cliff.yaml --ckpt data/poco_cliff.pt --output_folder out
To display the uncertainty values, shape categorization, or person id numbers as text overlay, you may add the following flags:
--display_uncertainty --categorize_shape --display_id
File outputs:
To download the output files from the Docker Container, use the Docker application to open the container and look within the corresponding folders for each model.
The output for deep_sort will be found in the root folder of deep_sort named output.mp4
The output for POCO will be found within
/POCO/out/ShootingInOTR_/<file_name>_results.mp4
For more information on this repository please contact the creator.
Content type
Image
Digest
sha256:a1fb75477…
Size
8.8 GB
Last updated
about 2 years ago
docker pull telepathboy/eeai_hybrid:hybrid_model