Semi-automatic Image Annotator

For labeling data to train object detection models, I could have used a tool like LabelImg, but I decided to write my own tool. First, I wanted to lean how to develop such a tool and second, I wanted to integrate a pre-trained model to suggest bounding boxes. This way, I can speed up the labeling process. As a pre-trained model, I use the YOLOv8 model from the ultralytics framework. In the following video, I show a demonstration of the tool on some images from the Dogs vs. Cats dataset. The tool is written in Python.

The code for this project can be found in my GitHub repository.