ATSO: Asynchronous Teacher-Student Optimization for Semi-Supervised Medical Image Segmentation
This work addresses a specific bottleneck in semi-supervised learning for medical image analysis, offering an incremental improvement over existing self-learning pipelines.
The paper tackled the problem of lazy learning in semi-supervised medical image segmentation by proposing ATSO, an asynchronous teacher-student optimization method, which achieved superior performance on two medical datasets and transferred well to natural image segmentation.
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.