Lightweight Python library for adding real-time 2D object tracking to any detector.
object-tracking norfair computer-vision code object-detection library

Norfair is a customizable lightweight Python library for real-time 2D object tracking. Using Norfair, you can add tracking capabilities to any detector with just a few lines of code.


  • Any detector expressing its detections as a series of (x, y) coordinates can be used with Norfair. This includes detectors performing object detection, pose estimation, and instance segmentation.
  • The function used to calculate the distance between tracked objects and detections is defined by the user, making the tracker extremely customizable. This function can make use of any extra information, such as appearance embeddings, which can heavily improve tracking performance.
  • Modular. It can easily be inserted into complex video processing pipelines to add tracking to existing projects. At the same time it is possible to build a video inference loop from scratch using just Norfair and a detector.
  • Fast. The only thing bounding inference speed will be the detection network feeding detections to Norfair.

Don't forget to tag @tryolabs in your comment, otherwise they may not be notified.

Authors community post
We are a Machine Learning consulting shop with an active R&D division, focused on Deep Learning, Computer Vision and NLP.
Share this project
Similar projects
DeepSORT: Deep Learning to Track Custom Objects in a Video
A look at deep learning based approached for object tracking.
Object tracking in 75 lines of code
Object tracking is straightforward conceptually. And if you have a good detector, simple methods can be pretty effective.
TAO: A Large-Scale Benchmark for Tracking Any Object
A diverse dataset for Tracking Any Object (TAO) consisting of 2,907 high resolution videos, captured in diverse environments, which are half a minute long ...
The project deals with tracking humans in a narrow hallway under different lighting conditions.
Top collections