Hyperspectral imaging plays a vital role in homeland security, defence, medical imaging, and many other industrial applications. In general, Hyperspectral Images (HSI) are noisy. They require more accurate machine lea...
详细信息
ISBN:
(数字)9781665434454
ISBN:
(纸本)9781665434454
Hyperspectral imaging plays a vital role in homeland security, defence, medical imaging, and many other industrial applications. In general, Hyperspectral Images (HSI) are noisy. They require more accurate machine learning methods for object detection/prediction/classification and, more importantly, more stable prediction/ classification accuracies in the presence of noise. This work focuses on improving Hyperspectral Image (HSI) prediction/classification in the presence of noise. It demonstrates that it can outperform deep learning (DL) state-of-the-art results with a modified 3D-DenseNet architecture. Specifically, we (a) propose a new model that we denote 3D-HSI-DenseNet that relatively boosts classification accuracy of the original version of then 3D-DenseNet, while generally improving the prediction stability of the network in the presence of noise (b) demonstrate that the proposed method has a faster training and testing time in comparison to the original DenseNet;(c) show that the proposed modified architecture of the DenseNet reduces the total number of the model parameters by nearly 1.5 times compared to the DenseNet architecture from the HSI classification literature;(d) experimentally show that Network in Network (NIN) can be used as a Dimensionality Reduction (DR) technique and it is as effective as PCA, without its limitations, for band reduction of HSI;(e) show the stability and robustness in classification performance of the proposed model trained with and without noise, by introducing different levels of noise into the testing data set.
This paper works towards an initial ontology of assessment techniques for building AI-enriched human-centered XR systems, denoted Intelligent Realities (IRs). Rather than connecting technologies, our work analyses the...
详细信息
The proceedings contain 6 papers. The special focus in this conference is on Secure and Resilient Digital Transformation of Healthcare. The topics include: Threat modeling Towards Resilience in Smart ICUs;charact...
ISBN:
(纸本)9783031558283
The proceedings contain 6 papers. The special focus in this conference is on Secure and Resilient Digital Transformation of Healthcare. The topics include: Threat modeling Towards Resilience in Smart ICUs;characterizing Privacy Risks in Healthcare IoT systems;data Analytics, Digital Transformation, and Cybersecurity Perspectives in Healthcare;methodology for Automating Attacking Agents in Cyber Range Training Platforms.
Satellite image processing techniques have become increasingly important for monitoring and analyzing the Earth's surface and its features. Water bodies in particular are of great interest due to their ecological,...
详细信息
Reliability prediction in automotive systems undoubted represents a substantial part of safety and customer satisfaction. a new graph-based probabilistic method and machine learning algorithm for the automotive system...
详细信息
Conceptual modeling is gaining an essential role in designing large-scale systems, especially as the complexity of the latter increases. The design tool for such systems is the development of ontologies for their subj...
详细信息
Remote photoplethysmography (rPPG) is a promising non-contact method for measuring heart rate (HR) and physiological signals. However, current deep learning approaches in this field primarily focus on extracting subtl...
详细信息
The investigation aims to evaluate the performance of machine learning techniques, particularly the XGBoost regression method, for stock price prediction with the help of technical indicators. The research targets the...
详细信息
In an unstable market, supply chain management (SCM) must use novel approaches to meet consumers' expectations better. Supply chains must become more adaptable to cope with the disturbance that calls for increased...
详细信息
Although predictive AI models have grown to dominate computational finance, they are often limited in their applications when it comes to studying interventions and explaining behavioral outcomes. Financial economics,...
详细信息
暂无评论