期刊
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
卷 61, 期 -, 页码 -出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TGRS.2023.3264235
关键词
Feature extraction; Transformers; Task analysis; Training; Image reconstruction; Convolutional neural networks; Kernel; Contrastive learning (CL); deep learning; hyperspectral image (HSI) classification; pretraining; Transformer
In this study, a spectral-spatial masked Transformer (SS-MTr) is explored for hyperspectral image (HSI) classification. A two-stage training strategy is utilized, where the Transformer is pretrained via reconstruction and then fine-tuned. Additionally, three SS-MTr-based methods are proposed to incorporate discriminative feature learning. Experimental results demonstrate that these methods achieve competitive performance compared to state-of-the-art HSI classification methods.
Recently, due to the powerful capability at modeling the long-range relationships, Transformer-based methods have been widely explored in many research areas, including hyperspectral image (HSI) classification. However, because of lots of trainable parameters and the lack of inductive bias, it is difficult to train a Transformer-based HSI classifier, especially when the number of training samples is limited. To address this issue, in this study, spectral-spatial masked Transformer (SS-MTr) is explored for HSI classification, which uses a two-stage training strategy. In the first stage, SS-MTr pretrains a vanilla Transformer via reconstruction from masked HSI inputs, which embeds the local inductive bias into the Transformer. In the second stage, the well pretrained Transformer is cooperated with a fully connected layer and is then fine-tuned for the HSI classification. Furthermore, in order to incorporate discriminative feature learning into the SS-MTr, three SS-MTr-based methods, including contrastive SS-MTr (C-SS-MTr), supervised SS-MTr (S-SS-MTr), and supervised C-SS-MTr (SC-SS-MTr), are proposed by adding extra branches for specific tasks in parallel with the existing reconstruction task. Specifically, the proposed C-SS-MTr adds a contrastive loss, which brings instance discriminability. Besides, the proposed S-SS-MTr builds an extra classification branch for embracing interclass discriminability and intraclass similarity. Moreover, the proposed SC-SS-MTr combines C-SS-MTr and S-SS-MTr for better generalization. The proposed SS-MTr, C-SS-MTr, S-SS-MTr, and SC-SS-MTr are tested on three popular hyperspectral datasets (i.e., Indian Pines, Pavia University, and Houston). The obtained results reveal that the proposed models achieve competitive results compared with the state-of-the-art HSI classification methods. The code is available at https://github.com/mengduanjinghua/SS-MTr.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据