Here is an example of how you can run 3 object detection algorithms with the same 8 lines of code in a few seconds.
We use face detection from Kornia, YOLOv5 object detection from Ultralytics and Detectron2 keypoints from Meta AI.
More info on our Github repo : https://github.com/Ikomia-dev/IkomiaApi
How do we do this ? We take existing Python code and wrap it as ready-to-use algorithms for Ikomia API (our Python library). With this method, it’s easy to integrate individual repos from researchers or labs. We also integrated awesome frameworks like OpenCV, Detectron2, OpenMMLab or Hugging Face so that anyone can develop with the best state-of-the-art algorithms in a single framework.
Ikomia’s purpose is to share scientific knowledge. We believe that state-of-the-art algorithms are the key to success. That’s why we are building this community to make research-based algorithms ready-to-use for developers.