PENERAPAN METODE CONVOLUTIONAL NEURAL NETWORK DENGAN ARSITEKTUR EFFICIENTNET-B0 UNTUK KLASIFIKASI KESEGARAN DAGING AYAM

The freshness of chicken meat is a key factor in ensuring food safety and consumer satisfaction. This study aims to classify the freshness of chicken meat into two categories, Fresh and Non-Fresh, using a Convolutional Neural Network (CNN) with the EfficientNet-B0 architecture. The model was trained in two scenarios: without data augmentation and with augmentation techniques (flip, rotation, and zoom). Evaluation was conducted using accuracy, precision, recall, and f1-score metrics. The results show that the augmented model achieved 95% accuracy and an f1-score of 0.95, while the non-augmented model achieved only 90% accuracy and an f1-score of 0.89. These findings indicate that data augmentation improves model performance in classifying chicken meat freshness.