CuriousAI/mean-teacher
A semi-supervised learning method for image classification using exponential moving average of model weights as teacher.
Not currently ranked — collecting fresh signals.
star history
Mean Teacher is a semi-supervised learning technique published at NIPS 2017 that improves image classification by training a student network and maintaining a teacher network whose weights are an exponential moving average of the student’s weights. The method adds a consistency cost between student and teacher outputs, enabling effective learning from limited labeled data. It achieved state-of-the-art results on ImageNet and CIFAR-10 with only a fraction of labeled examples.
Frequently asked
- What is CuriousAI/mean-teacher?
- A semi-supervised learning method for image classification using exponential moving average of model weights as teacher.
- Is mean-teacher open source?
- Yes — CuriousAI/mean-teacher is an open-source project tracked on heatdrop.
- What language is mean-teacher written in?
- CuriousAI/mean-teacher is primarily written in Python.
- How popular is mean-teacher?
- CuriousAI/mean-teacher has 1.7k stars on GitHub.
- Where can I find mean-teacher?
- CuriousAI/mean-teacher is on GitHub at https://github.com/CuriousAI/mean-teacher.