Leaf Spectral Reflectance Prediction Using Multi-Head Attention Neural Networks
Researchers have developed a new multi-head attention neural network to predict leaf spectral reflectance in grapevines. This model, trained on a specific dataset of 16 leaf traits, achieved a high coefficient of determination (R^2) of 0.84 and a normalized root mean squared error (NRMSE) of 1.52 percent. It demonstrated superior accuracy compared to the traditional PROSPECT-PRO model, particularly in the NIR and SWIR regions, offering a more robust framework for remote sensing applications in agriculture. AI
IMPACT Enhances precision agriculture and plant science by improving spectral reflectance modeling for crop management.