This is a binary image classification project that uses a convolutional neural network (CNN) to detect whether an image is a "hotdog" or "not hotdog." The project utilizes transfer learning to fine-tune several pre-trained models, including AlexNet, VGG19, ResNet50, and Inception V3, for image classification.
After comparing the results of these models, the Inception ResNet V2 model was found to produce the most outstanding results, and thus was used as the final model in this project.
The dataset comprises 4,642 images in total, split evenly between "hotdog" and "not hotdog" categories, with 2,121 images for training and 200 images for testing.
The project uses a pre-trained Inception ResNet V2 model as the base model, with a dense layer of 2 units added for binary classification. The model was fine-tuned for 50 epochs on the training set with a batch size of 32.
The trained model achieved 94.25% accuarcy and 0.1438 loss on the test set.
The project was developed using Google Colab, a free cloud-based platform that provides easy access to GPUs and TPUs, making it ideal for machine learning projects.
