What Will You Learn in The Project Computer vision project?
- Loading the image data with ImageDataGenerator class
- Developing CNN model and compiling and fitting it on the train set
- How to perform Data Augmentation?
- How to do transfer learning with pre-trained models such as Vgg16?
Pre-Requisite
- Woking knowledge of Keras library
- Understanding of different convolution operations i.e., Filter, kernels, padding, pooling
Tools Used
- Google Colab for training the model
- Matplotlib library for visualizing the loss & accuracy of the model
- Keras[tensorflow]
Tasks To Be Performed
We will be performing the following tasks as part of this project:
Task-1: Import the dataset and divide it into a train/test/validation split.
Task-2: Develop a fully connected vanilla(base) CNN model
Task-3: Compile the model and do training with Early Stopping
Task-4: Validate the base model on the test set
Task-5: Perform data augmentation on train set
Task-6: Train and evaluate the updated model with augmented data
Task-7: Download the vgg16 model and freeze it’s all the weights
Task-8: Create a new model with vgg16
Task-9: Fit and evaluate the updated vgg model
Task-10: Save the best model into Keras H5 format