Exploring Deep Learning for Comprehensive Dog Breed Identification

The breed of a dog can be identified from an image using computer vision and deep learning techniques. With the use of user-submitted dog images, the model will identify the breed of the dog by choosing from among the 70 breeds included in the dataset. Inception, Xception, InceptionResNetV2, and MobileNetV2 are a few of the deep learning algorithms that are used in the proposed work to determine dog breeds. The Xception, Inception, InceptionResNetV2, and MobileNetV2 models exhibit high accuracy rates in the current system, with $\mathbf{9 1. 6 7 \%}, \mathbf{9 0. 9 2 \%}$, and $\mathbf{9 0. 0 3 \%}$, respectively. The usefulness of these algorithms in identifying breeds is evaluated through comparison. The outstanding performance of these algorithms especially Xception, which attained the highest accuracy is highlighted in this research. These models’ efficacy shows how deep learning techniques have made tremendous progress in dog breed recognition.