microsoft/SoftTeacher
A semi-supervised object detection system that uses a teacher-student training paradigm to leverage unlabeled images.
Not currently ranked — collecting fresh signals.
star history
SoftTeacher is an end-to-end semi-supervised object detection framework that jointly trains a student detector and a teacher model. The teacher generates soft pseudo-labels for unlabeled images, which the student uses for training. This approach improves detection accuracy by utilizing unlabeled data alongside labeled data without requiring additional annotation costs.
Frequently asked
- What is microsoft/SoftTeacher?
- A semi-supervised object detection system that uses a teacher-student training paradigm to leverage unlabeled images.
- Is SoftTeacher open source?
- Yes — microsoft/SoftTeacher is open source, released under the MIT license.
- What language is SoftTeacher written in?
- microsoft/SoftTeacher is primarily written in Python.
- How popular is SoftTeacher?
- microsoft/SoftTeacher has 926 stars on GitHub.
- Where can I find SoftTeacher?
- microsoft/SoftTeacher is on GitHub at https://github.com/microsoft/SoftTeacher.