Tech Xplore

Stress-testing AI vision systems: Rethinking how adversarial images are generated

Deep neural networks (DNNs) have become a cornerstone of modern AI technology, driving a thriving field of research in ...
Frontiers

A privacy-preserving, on-board satellite image classification technique incorporating homomorphic encryption and transfer learning

Satellite image classification is an important and challenging task in the modern technological age. Satellites can capture images of danger-prone areas with very little effort. However, the size and ...
Frontiers

VAE deep learning model with domain adaptation, transfer learning and harmonization for diagnostic classification from multi-site neuroimaging data

In large public multi-site fMRI datasets, the sample characteristics, data acquisition methods, and MRI scanner models vary across sites and datasets. This non-neural variability obscures neural ...
IEEE

Open Set Cross-Domain Hyperspectral Image Classification Based on Critical Reflective Learning Network

Abstract: Limited by the lack of supervisory information of unknown classes, the existing open set cross-domain hyperspectral image (HSI) classification methods often rely on threshold-based methods ...
Law

Training Alone Is Not Enough: Lessons From 'Recentive' and USPTO AI Examples on Patent Eligible Machine-Learning Claims

The Recentive decision exemplifies the Federal Circuit’s skepticism toward claims that dress up longstanding business problems in machine-learning garb, while the USPTO’s examples confirm that ...
C&EN

Transfer-Learning Deep Raman Models Using Semiempirical Quantum Chemistry

Creative Commons (CC): This is a Creative Commons license. Attribution (BY): Credit must be given to the creator. Biophotonic technologies such as Raman spectroscopy are powerful tools for obtaining ...
IEEE

Incremental Classification of Cross-Scene Hyperspectral Images Based on Dual Constraints and Knowledge Transfer

Abstract: With the rapid advancement of hyperspectral imaging technology, there has been a dramatic surge in the volume of hyperspectral data, presenting unprecedented challenges for the incremental ...