Fig 1.
Flowchart for the proposed GNViT–Groundnut crop pest classification using Vision Transformer (ViT) model.
Table 1.
The generalized pseudo code for the GNViT model.
Fig 2.
Showing a single image becomes 256 image patches—Republished from [45] under a CC BY license, with permission from [MDPI], original copyright [2022].
Fig 3.
An overview of the proposed GNViT framework for identifying groundnut crop pests—Republished from [45, 46] under a CC BY license, with permission from [MDPI], original copyright [2022, 2023].
The (*) notation appended to the patch + position embedding signifies the presence of a class token, which serves as a pivotal element encapsulating comprehensive image information within the sequence of patch embeddings.
Table 2.
GNViT learning-model-training parameters.
Fig 4.
GNViT model performance on groundnut pest classification accuracy for 10 epochs with Augmentation.
Table 3.
GNViT model training results.
Fig 5.
GNViT model performance on groundnut pest classification accuracy for 10 epochs without Augmentation.
Table 4.
Comparison of accuracy between GNViT model and existing models.