Attention-enhanced VGG-16 architecture for precision weed detection in wheat fields

This research details five state of the art convolutional neural-network designs that use deep learning to classify wheat weeds. The dataset consists of greyscale photos taken in agricultural fields, enhanced with RGB lighting and cropped to 256×256×3 pixels. A VGG16-based model with further Conv2D layers and spatial attention is one of the five models suggested. Another is an architecture inspired by ResNet50 that uses residual blocks and global average pooling. For accurate wheat-weed classification, the study found that VGG16 is the best option because to its high level of generalizability and precision.