Classification of Beef and Pork Using a Hybrid Model of ResNet-50 and Support Vector Machine (SVM)

Some people manipulate sales in marketplaces and other retail settings by combining beef and pork since the prices are so high. In addition to educating the public about these distinctions, this research aims to develop a technological solution for recognizing and differentiating between pork and beef. The proposed hybrid model, a combination of a Resnet-50 and a Support Vector Machine (SVM), is introduced for the classification of Beef and Pork Meat. In this hybrid model, the Resnet-50 functions as a powerful feature extractor, then utilizing its inherent ability to automatically capture distinctive features from diverse and highly specific meat image datasets. The SVM, serving as the binary classifier, effectively utilizes the extracted features for precise classification. The hybrid model achieves an outstanding accuracy of 100%, surpassing the performance of individual classifiers, with Resnet-50 achieving 97% accuracy and Resnet-50 achieving 97% obtained from the Hybrid model by gaining the best parameter C is 0,1 and the Kernel is linear. This remarkable outcome signifies the synergistic effectiveness of combining Resnet-50 and SVM.