Fruit ripeness identification using transformers

Abstract Pattern classification has always been essential in computer vision. Transformer paradigm having attention mechanism with global receptive field vision improves the efficiency and effectiveness of visual object detection recognition. The primary purpose this article is to achieve accurate ripeness various types fruits. We create fruit datasets train, test, evaluate multiple models. Transformers are fundamentally composed encoding decoding procedures. encoder stack blocks, like convolutional neural networks (CNN or ConvNet). Vision (ViT), Swin Transformer, multilayer perceptron (MLP) considered paper. examine advantages these three models for accurately analyzing ripeness. find that achieves more significant outcomes than ViT both pears apples from our dataset.