Multiple Object Tracking Systems


Date:  Thursday, May 23

Start Time: 11:25 am

End Time: 12:30 pm

Object tracking is an essential capability in many computer vision systems, including applications in fields such as traffic control, self-driving vehicles, sports and more. In this session, we’ll walk through the construction of a typical multiple object tracking (MOT) algorithm step by step. At each step, we’ll identify key challenges and explore design choices (for example, detection-based vs. detection-free approaches and online vs. offline tracking). We’ll discuss available off-the-shelf MOT algorithms and open-source libraries. We’ll also identify areas where current MOT algorithms fall short. And we’ll introduce metrics and benchmarks commonly used to evaluate MOT solutions.

Track

Session Speakers

  • With over ten years of experience in software development and machine learning, Javier Berneche is a seasoned Senior Machine Learning Engineer at Tryolabs. He has a track record of success in diverse industries such as retail, medicine and disaster prevention, and has expertise in tackling challenging computer vision problems such as classification, object detection and keypoint detection. In addition to his expertise in computer vision, Javier has gained valuable experience in natural language processing and tabular data processing. He has also led multiple machine learning projects, managing aspects including labeling, model training, MLOps, explainability and model evaluation. As the head of the computer vision team at Tryolabs, Javier is responsible for the development and maintenance of Norfair, an open-source library for multiple object tracking.


Share

Multiple Object Tracking Systems