AINeutralarXiv – CS AI · 10h ago5/10
🧠
Few-Shot Hyperspectral Aphid Detection via FastGAN Synthetic Data Generation, Transformer-Based Classification and Explainable AI
Researchers developed a FastGAN-based synthetic data generation method to augment limited hyperspectral imaging datasets for detecting aphid infestations in crops, achieving superior classification results with Vision Transformer models. The approach demonstrates how generative AI and transformer architectures can overcome data scarcity challenges in agricultural pest detection, enabling more efficient and accurate crop monitoring.