Content Based Image Recognition

Content Based Image Recognition model built using Keras which uses Sequential Model which achieves training accuracy of 97.62% and Validation Accuracy of 95.25% on data which consists of almost 1000 images of 5 classes viz: Aeroplane, Bike, Boats, Cars, Dogs. This accuracy is not achieved by using Image Data Generator which is implemented in the code. The second code: fashionMNIST.ipynb also built using Keras and uses Sequential model which achieves Training Accuracy of 94.49% and Test Accuracy of 93.55%.