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 2502.17879cs.CV
收藏

双分类头自训练网络用于跨场景高光谱图像分类

Dual Classification Head Self-training Network for Cross-scene Hyperspectral Image Classification

University of Wisconsin-Madison(威斯康星大学麦迪逊分校);Technical University of Munich(慕尼黑工业大学);The University of Hong Kong(香港大学)

Rong Liu, Junye Liang, Jiaqi Yang, Jiang He, Peng Zhu

AI总结:

针对跨场景高光谱图像分类中的域差距问题,提出双分类头自训练网络DHSNet,通过类别级特征对齐和中心特征注意力机制,在三个数据集上显著优于现有方法。

摘要:

由于高光谱图像(HSIs)获取标记数据困难,跨场景分类已成为遥感领域广泛采用的方法。该方法利用源域(SD)的标记数据和目标域(TD)的未标记数据训练模型,随后在TD上进行推理。然而,同一物体在SD和TD之间的反射光谱差异,以及同一土地覆盖类别特征分布的差异,对跨场景分类性能构成了重大挑战。为解决此问题,我们提出了一种双分类头自训练网络(DHSNet)。该方法跨域对齐类别级特征,确保训练后的分类器能够准确分类TD中不同类别的数据。我们首次在跨场景HSI分类领域引入双分类头自训练策略。所提出的方法缓解了域差距,同时防止模型中错误伪标签的累积。此外,我们引入了一种新颖的中心特征注意力机制,以增强模型跨域学习场景不变特征的能力。在三个跨场景HSI数据集上的实验结果表明,所提出的DHSNet显著优于其他最先进的方法。DHSNet的代码将在https://github.com/liurongwhm提供。

英文摘要:

Due to the difficulty of obtaining labeled data for hyperspectral images (HSIs), cross-scene classification has emerged as a widely adopted approach in the remote sensing community. It involves training a model using labeled data from a source domain (SD) and unlabeled data from a target domain (TD), followed by inferencing on the TD. However, variations in the reflectance spectrum of the same object between the SD and the TD, as well as differences in the feature distribution of the same land cover class, pose significant challenges to the performance of cross-scene classification. To address this issue, we propose a dual classification head self-training network (DHSNet). This method aligns class-wise features across domains, ensuring that the trained classifier can accurately classify TD data of different classes. We introduce a dual classification head self-training strategy for the first time in the cross-scene HSI classification field. The proposed approach mitigates domain gap while preventing the accumulation of incorrect pseudo-labels in the model. Additionally, we incorporate a novel central feature attention mechanism to enhance the model's capacity to learn scene-invariant features across domains. Experimental results on three cross-scene HSI datasets demonstrate that the proposed DHSNET significantly outperforms other state-of-the-art approaches. The code for DHSNet will be available at https://github.com/liurongwhm.

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。

↑