Creating a Sequential CNN model to classify images of various datasets and comparing the results to pretrained models (VGG16 and Inception V3). A dashboard design for the CNN model for the prediction
This project is on classification of images on two different datasets containing atotal of 8992 images of which the model used 75 percent dataset to train and 25 percent dataset to test the results. The accuracy obtained by using a basic Sequential CNN model is almost 73 percent. The results generated by the pretrained models of Oxford (VGG16) and Google (Inception V3)is more accurate by a difference of 10 percent than our custom model.
Please note that dashboard will be hosted on 127.0.0.1:8050
