Binary Image Classification Using VGG Architecture

Binary Image Classification Using VGG Architecture

Mar, 2024 - Apr, 2024 (Advisor – Prof. Prof. Nipun Batra, IIT Gandhinagar) | Project Link

• Implemented VGG1 and VGG3 architectures for Convolutional Neural Networks and tested their performance on the classification of two image classes.
• Performed Augmentation over the dataset and trained the VGG3 model over this augmented dataset.
• Performed Transfer Learning over the pre-trained VGG16 models while both freezing and un-freezing the fully connected layers.
• Compared the performance of all these models by the gradual increase in classification accuracy over the testing images, changes in training and validation losses using Tensorboard.