arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.18536cs.CV
收藏

大型掩码自编码器预训练对下游地球观测任务的有效性如何?

How Effective is Pre-training of Large Masked Autoencoders for Downstream Earth Observation Tasks?

University of Luxembourg(卢森堡大学);Hydrosat(Hydrosat公司)

Jose Sosa, Mohamed Aloulou, Danila Rukhovich, Rim Sleimi, Boonyarit Changaival, Anis Kacem, Djamila Aouada

AI总结:

本研究调查了基于ViT的MAE预训练在下游地球观测任务中的有效性,发现预训练在任务与预训练相似时有益,但在分割或分类任务中从头训练同样或更有效。

摘要:

自监督预训练已被证明对许多计算机视觉任务非常有效,尤其是在标注数据稀缺的情况下。在地球观测(EO)背景下,基础模型和各种其他基于 Vision Transformer (ViT) 的方法已成功应用于下游任务的迁移学习。然而,目前尚不清楚在何种条件下,预训练模型相比从头训练能提供显著优势。在本研究中,我们调查了基于 ViT 的 Masked Autoencoders (MAE) 预训练对下游 EO 任务的有效性,重点关注重建、分割和分类。我们考虑了两个大型基于 ViT 的 MAE 预训练模型:一个基础模型(Prithvi)和 SatMAE。我们在基于重建和分割的下游任务上评估了 Prithvi,对于 SatMAE,我们评估了其在分类下游任务上的性能。我们的研究结果表明,当微调任务与预训练任务(例如重建)高度相似时,预训练尤其有益。相比之下,对于分割或分类等任务,从头训练并配合特定的超参数调整被证明同样有效甚至更有效。

英文摘要:

Self-supervised pre-training has proven highly effective for many computer vision tasks, particularly when labelled data are scarce. In the context of Earth Observation (EO), foundation models and various other Vision Transformer (ViT)-based approaches have been successfully applied for transfer learning to downstream tasks. However, it remains unclear under which conditions pre-trained models offer significant advantages over training from scratch. In this study, we investigate the effectiveness of pre-training ViT-based Masked Autoencoders (MAE) for downstream EO tasks, focusing on reconstruction, segmentation, and classification. We consider two large ViT-based MAE pre-trained models: a foundation model (Prithvi) and SatMAE. We evaluate Prithvi on reconstruction and segmentation-based downstream tasks, and for SatMAE we assess its performance on a classification downstream task. Our findings suggest that pre-training is particularly beneficial when the fine-tuning task closely resembles the pre-training task, e.g. reconstruction. In contrast, for tasks such as segmentation or classification, training from scratch with specific hyperparameter adjustments proved to be equally or more effective.

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。

↑