Comparative Analysis of Pre-trained CNN Models for Image Classification of Emergency Vehicles

This paper presents a comprehensive study on image classification, focusing specifically on the evaluation of six pre-trained CNN models in the context of emergency vehicle classification. The models examined in this research are VGG19, VGG16, MobileNetV3Large, MobileNetV3Small, MobileNetV2, and MobileNetV1. We followed a systematic research methodology involving dataset preparation, model architecture modification, layer operation restriction, and model compilation optimization. Through extensive experimentation, the performance of each model is analyzed, considering factors such as accuracy, loss and training time. The findings shed light on the strengths and limitations of each model, emphasizing the importance of selecting an appropriate pre-trained CNN model for image classification tasks. Overall, this article provides a comprehensive overview of image classification, highlighting the crucial role of pre-trained CNN models in achieving accurate results for emergency vehicle classification.