Harnessing Deep Learning for Cat Breed Identification: A Visual Recognition Strategy

This research aims to accurately recognize and identify various cat breeds, a significant endeavor for animal lovers. Cats are cherished pets worldwide, yet distinguishing between different breeds can be quite challenging. This study investigates the application of image processing techniques to classify cat breeds, utilizing several convolutional neural network (CNN) models to achieve optimal accuracy.In our experiments, we evaluated seven CNN models: DenseNet121, ResNet50V2, InceptionV3, VGGNet16, VGGNet19, EfficientNetB0, and MobileNetV2. These models were assessed using various metrics, including accuracy, precision, recall, and F1 score. For the DenseNet121, InceptionV3, and ResNet50V2 models, images were resized to 299x299 pixels, whereas for VGGNet16, VGGNet19, MobileNetV2, and EfficientNetB0, images were resized to 224x224 pixels.The results indicate that EfficientNetB0, VGGNet16, and VGGNet19 outperformed the other models, achieving accuracy metrics of 76% or higher. Future research will concentrate on enhancing data augmentation and fitting techniques to further improve the performance of CNN models.