arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2403.13341cs.CVcs.AI
收藏

FissionFusion:用于医学图像分析的快速几何生成与分层融合

FissionFusion: Fast Geometric Generation and Hierarchical Souping for Medical Image Analysis

Mohamed bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学)

Santosh Sanjeev, Nuren Zhaksylyk, Ibrahim Almakky, Anees Ur Rehman Hashmi, Mohammad Areeb Qazi, Mohammad Yaqub

AI总结:

提出FissionFusion方法,通过分层合并与循环学习率高效生成模型,显著提升医学图像分析在ID和OOD数据集上的性能。

摘要:

标注良好的医学数据集的稀缺性需要利用来自更广泛数据集(如ImageNet)或预训练模型(如CLIP)的迁移学习。模型汤(model soups)通过平均多个微调模型,旨在提升In-Domain (ID)任务的性能并增强对Out-of-Distribution (OOD)数据集的鲁棒性。然而,将这些方法应用于医学成像领域面临挑战并导致次优性能。这主要是由于数据复杂性(如异质性、域偏移、类别不平衡以及训练和测试阶段之间的分布偏移)导致的误差表面特征差异。为了解决这个问题,我们提出了一种分层合并方法,该方法基于模型的超参数配置,在不同层级对模型进行局部和全局聚合。此外,为了减轻在超参数搜索中训练大量模型的需求,我们引入了一种计算高效的方法,使用循环学习率调度器在权重空间中生成多个模型以进行聚合。我们的方法在多个数据集上比模型汤方法展现出显著改进(在HAM10000和CheXpert数据集上约有6%的提升),同时保持了较低的模型生成和选择的计算成本。此外,我们在OOD数据集上取得了优于模型汤的结果。代码可在https://github.com/BioMedIA-MBZUAI/FissionFusion获取。

英文摘要:

The scarcity of well-annotated medical datasets requires leveraging transfer learning from broader datasets like ImageNet or pre-trained models like CLIP. Model soups averages multiple fine-tuned models aiming to improve performance on In-Domain (ID) tasks and enhance robustness against Out-of-Distribution (OOD) datasets. However, applying these methods to the medical imaging domain faces challenges and results in suboptimal performance. This is primarily due to differences in error surface characteristics that stem from data complexities such as heterogeneity, domain shift, class imbalance, and distributional shifts between training and testing phases. To address this issue, we propose a hierarchical merging approach that involves local and global aggregation of models at various levels based on models' hyperparameter configurations. Furthermore, to alleviate the need for training a large number of models in the hyperparameter search, we introduce a computationally efficient method using a cyclical learning rate scheduler to produce multiple models for aggregation in the weight space. Our method demonstrates significant improvements over the model souping approach across multiple datasets (around 6% gain in HAM10000 and CheXpert datasets) while maintaining low computational costs for model generation and selection. Moreover, we achieve better results on OOD datasets than model soups. The code is available at https://github.com/BioMedIA-MBZUAI/FissionFusion.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑