Binder is a free service where the webcam will not work, and performance on the video will not be good. For the best performance, install the notebooks locally.
This notebook shows a person tracking scenario: it reads frames from an input video sequence, detects people in the frames, uniquely identifies each one of them and tracks all of them until they leave the frame.
This tutorial uses the Deep SORT algorithm to perform object tracking. person detection model is deployed to detect the person in each frame of the video, and reidentification model is used to output embedding vector to match a pair of images of a person by the cosine distance.
This is a self-contained example that relies solely on its own code.
We recommend running the notebook in a virtual environment. You only need a Jupyter server to start.
For details, please refer to Installation Guide.