ATSO: Asynchronous Teacher-Student Optimization for Semi-Supervised Medical Image Segmentation

TL;DR

This paper introduces ATSO, an asynchronous teacher-student method for semi-supervised medical image segmentation that improves learning from pseudo labels by alternating data subsets, demonstrating superior performance and transferability.

Contribution

The paper proposes ATSO, a novel asynchronous teacher-student framework that addresses lazy learning in semi-supervised segmentation by partitioning data and alternating updates.

Findings

ATSO outperforms existing semi-supervised methods on medical segmentation datasets.

The method achieves state-of-the-art results in various semi-supervised settings.

ATSO effectively transfers to natural image segmentation tasks.

Abstract

In medical image analysis, semi-supervised learning is an effective method to extract knowledge from a small amount of labeled data and a large amount of unlabeled data. This paper focuses on a popular pipeline known as self learning, and points out a weakness named lazy learning that refers to the difficulty for a model to learn from the pseudo labels generated by itself. To alleviate this issue, we propose ATSO, an asynchronous version of teacher-student optimization. ATSO partitions the unlabeled data into two subsets and alternately uses one subset to fine-tune the model and updates the label on the other subset. We evaluate ATSO on two popular medical image segmentation datasets and show its superior performance in various semi-supervised settings. With slight modification, ATSO transfers well to natural image segmentation for autonomous driving data.