Reliable Explainability of Deep Learning Spatial-Spectral Classifiers for Improved Semantic Segmentation in Autonomous Driving
This work addresses the need for reliable explainability in safety-critical autonomous driving systems, though it appears incremental as it builds on existing saliency methods.
The paper tackled the problem of unreliable explainability in deep learning spatial-spectral classifiers for semantic segmentation in autonomous driving by proposing an alternative approach using activations and weights from DNN layers, resulting in improved assessment of hyperspectral imagery performance and enhanced robustness through spectral signature normalization.
Integrating hyperspectral imagery (HSI) with deep neural networks (DNNs) can strengthen the accuracy of intelligent vision systems by combining spectral and spatial information, which is useful for tasks like semantic segmentation in autonomous driving. To advance research in such safety-critical systems, determining the precise contribution of spectral information to complex DNNs' output is needed. To address this, several saliency methods, such as class activation maps (CAM), have been proposed primarily for image classification. However, recent studies have raised concerns regarding their reliability. In this paper, we address their limitations and propose an alternative approach by leveraging the data provided by activations and weights from relevant DNN layers to better capture the relationship between input features and predictions. The study aims to assess the superior performance of HSI compared to 3-channel and single-channel DNNs. We also address the influence of spectral signature normalization for enhancing DNN robustness in real-world driving conditions.