Day #228 - OpenVino and pedestrian_tracker_demo

1. Download Demo Projects from link
2. Follow instructions in link to build demos

3. Execute commands

mkdir build

cd build

cmake -DCMAKE_BUILD_TYPE=Release <path_to_inference_engine_demos_directory>

make (Please check below)

4. pedestrian_tracker_demo - Documentation 

How it works

Step 1 -  Primary detection network for finding pedestrians

Step 2 -  Inference of the first network and makes reidentification of the pedestrians

Command to Run pedestrian_tracker

./pedestrian_tracker_demo -i <path_video_file> \

                          -m_det <path_person-detection-retail-0013>/person-detection-retail-0013.xml \

                          -m_reid <path_person-reidentification-retail-0031>/person-reidentification-retail-0031.xml \

                          -d_det GPU

5. Actual Command

./pedestrian_tracker_demo -i /home/ubuntu/code/smarthub_915am_cut.mp4.mp4 -m_det /opt/intel/computer_vision_sdk/deployment_tools/intel_models/person-detection-retail-0013/FP32/person-detection-retail-0013.xml -m_reid /opt/intel/computer_vision_sdk/deployment_tools/intel_models/person-reidentification-retail-0031/FP32/person-reidentification-retail-0031.xml -d_det CPU

Finally I was able to run through this. It is interesting they have two models in real time - Detection and Re-Identification. It will be interesting to know how this model has been trained, the counting of people re-entering, the inferences that is done through this. There must be a base data for training which decides detection, Also the re-id logic of comparison between the detected and similar one. 

Not Needed

1. Goto Download Directory ./downloader.py --all

Happy Mastering DL!!!