Vision Transformer Approach for Vegetables Recognition

The use of Vision Transformer to solve computer vision issues, particularly Image Classification, is a recent trend. Smart agriculture is one of the objectives of Industry 4.0 in Indonesia, which is currently vigorously advancing the agricultural sector. Vegetable classification is one example of a difficulty that can be encountered in the approach of smart agriculture. As a result of this possible issue, one solution is to use Vision Transformer to construct deep learning models. In this study, we use the Vision Transformer to tackle the problem of vegetable classification with an input size of 32x32 and a total patch size of 64. The model is constructed and trained with, and it has a 98% accuracy. Furthermore, the model employs several measures to evaluate the performance of the developed model. These findings indicate a promising performance for image classification problems, particularly with vegetables recognition.