Iterative Bounding Box Annotation for Object Detection

This paper proposes an annotation framework that uses an incremental learning approach on a small batch of manually labeled images, trains a detection model, uses freshly trained model to propose bounding boxes on a batch of unlabeled images, and requests the annotator do the correction on possible incorrect bounding boxes or labels proposals.

Cite this article as:

@INPROCEEDINGS{9412956,
  author={Adhikari, Bishwo and Huttunen, Heikki},
  booktitle={2020 25th International Conference on Pattern Recognition (ICPR)}, 
  title={Iterative Bounding Box Annotation for Object Detection}, 
  year={2021},
  pages={4040-4046},
  doi={10.1109/ICPR48806.2021.9412956}}