Dog Care: A Comprehensive Dog Breed Classification, Recommendation and Disease Prediction using Deep Learning
Dr. S Manjula, Chaitra Ishwar Hiremath, Sumukh D N, Branda Hijam, Sushil Kumar M S
Dept of Computer Science and Engineering
JSS Science and Technology University
Mysuru, Karnataka, India
Abstract---This research introduces a comprehensive system that combines dog breed classification, personalized breed recommendation, and disease prediction using deep learning and machine learning techniques. The first component utilizes a Convolutional Neural Network (CNN) model to accurately classify dog breeds from images, addressing the challenges of visual similarity between breeds and varying image conditions. The model is trained on a diverse dataset, achieving strong generalization across multiple categories. Building on this, a breed recommendation module is developed using a hybrid content-based approach that considers user preferences, lifestyle attributes, and breed characteristics to suggest the most suitable dog breeds for potential owners or adopters. The final component of the system focuses on disease prediction, leveraging breed-specific data and user-provided symptoms to identify potential health conditions using supervised machine learning models. This module aims to assist in early detection and preventive care, offering support to veterinarians and pet owners alike. Extensive experimentation and evaluation show that the integrated system performs reliably across all three tasks. By unifying these functionalities, the proposed framework not only enhances pet care and management but also promotes informed decision-making and responsible dog ownership through the application of intelligent, data-driven methods.
Keywords: Convolutional Neural Networks (CNN), Transfer Learning, Image-Based Breed Classification, Hybrid Recommendation Systems, Content-Based Filtering, Supervised Machine Learning, Multiclass Image Classification, Breed-Specific Disease Prediction, Feature Extraction, Model Evaluation Metrics (Accuracy, Precision, Recall, F1-Score).
