Progressive Growing of Patch Size: Resource-Efficient Curriculum Learning for Dense Prediction Tasks

TL;DR

This paper presents a resource-efficient curriculum learning method that progressively increases patch size during training, leading to faster convergence and improved segmentation accuracy in medical image analysis.

Contribution

It introduces the first successful use of a patch size curriculum in computer vision, reducing training costs while enhancing performance on dense prediction tasks.

Findings

Reduces training runtime by approximately 50%

Outperforms standard nnU-Net in 7 out of 10 tasks

Decreases computational costs and CO2 emissions

Abstract

In this work, we introduce Progressive Growing of Patch Size, a resource-efficient implicit curriculum learning approach for dense prediction tasks. Our curriculum approach is defined by growing the patch size during model training, which gradually increases the task's difficulty. We integrated our curriculum into the nnU-Net framework and evaluated the methodology on all 10 tasks of the Medical Segmentation Decathlon. With our approach, we are able to substantially reduce runtime, computational costs, and CO2 emissions of network training compared to classical constant patch size training. In our experiments, the curriculum approach resulted in improved convergence. We are able to outperform standard nnU-Net training, which is trained with constant patch size, in terms of Dice Score on 7 out of 10 MSD tasks while only spending roughly 50% of the original training runtime. To the best of our knowledge, our Progressive Growing of Patch Size is the first successful employment of a sample-length curriculum in the form of patch size in the field of computer vision. Our code is publicly available at https://github.com/compai-lab/2024-miccai-fischer.