A Biologically Interpretable Two-Stage Deep Neural Network (BIT-DNN) for Vegetation Recognition From Hyperspectral Imagery
A Biologically Interpretable Two-Stage Deep Neural Network (BIT-DNN) for Vegetation Recognition From Hyperspectral Imagery
复制标题
一种用于从高光谱图像中识别植被的具有生物学可解释性的两阶段深度神经网络（BIT - DNN）
DOI：
10.1109/tgrs.2021.3058782
复制
发表时间：
2021-02
影响因子：
8.2
通讯作者：
Yue Shi;Liangxiu Han;Wenjiang Huang;Sheng Chang;Yingying Dong;D. Dancey;Lianghao Han
中科院分区：
文献类型：
--
作者：
Yue Shi;Liangxiu Han;Wenjiang Huang;Sheng Chang;Yingying Dong;D. Dancey;Lianghao Han
基于光谱空间的深度学习模型最近被证明在高光谱图像（HSI）分类中是有效的，用于各种地球监测应用，如土地覆盖分类和农业监测。然而，由于“黑箱”模型表示的性质，如何解释和解释的学习过程和模型的决定，特别是植被分类，仍然是一个开放的挑战。这项研究提出了一种新的可解释的深度学习模型-生物可解释的两阶段深度神经网络（BIT-DNN），通过结合先验知识（即，生物物理和生物化学属性及其目标实体的层次结构）的光谱空间特征转换到所提出的框架，能够实现高精度和基于HSI的分类任务的可解释性。该模型引入了两阶段的特征学习过程：第一阶段，增强的可解释特征块提取与目标实体的生物物理和生物化学属性相关的低级别光谱特征;在第二阶段，可解释的封装块提取并封装高级联合频谱，代表这些目标实体的生物物理和生物化学属性的分级结构的空间特征，这为模型提供了改进的分类性能和内在的可解释性，同时降低了计算复杂度。我们已经使用四个真实的HSI数据集对四个单独的任务（即，植物物种分类、土地覆盖分类、城市场景识别和作物病害识别任务）。该模型与五种最先进的深度学习模型进行了比较。结果表明，该模型在分类精度和模型可解释性方面具有竞争优势，特别是在植被分类方面。
Spectral–spatial-based deep learning models have recently proven to be effective in hyper-spectral image (HSI) classification for various earth monitoring applications such as land cover classification and agricultural monitoring. However, due to the nature of “black-box” model representation, how to explain and interpret the learning process and the model decision, especially for vegetation classification, remains an open challenge. This study proposes a novel interpretable deep learning model—a biologically interpretable two-stage deep neural network (BIT-DNN), by incorporating the prior-knowledge (i.e., biophysical and biochemical attributes and their hierarchical structures of target entities)-based spectral–spatial feature transformation into the proposed framework, capable of achieving both high accuracy and interpretability on HSI-based classification tasks. The proposed model introduces a two-stage feature learning process: in the first stage, an enhanced interpretable feature block extracts the low-level spectral features associated with the biophysical and biochemical attributes of target entities; and in the second stage, an interpretable capsule block extracts and encapsulates the high-level joint spectral–spatial features representing the hierarchical structure of biophysical and biochemical attributes of these target entities, which provides the model an improved performance on classification and intrinsic interpretability with reduced computational complexity. We have tested and evaluated the model using four real HSI data sets for four separate tasks (i.e., plant species classification, land cover classification, urban scene recognition, and crop disease recognition tasks). The proposed model has been compared with five state-of-the-art deep learning models. The results demonstrate that the proposed model has competitive advantages in terms of both classification accuracy and model interpretability, especially for vegetation classification.