Categorical Image recognition using deep learning models:
CNN Model with Regularization CNN Model with Image Augmentation Leveraging Transfer Learning with Pre-trained CNN Models VGG-16 model Pre-trained CNN model as a Feature Extractor with Image Augmentation
The 'Auto-tag Images of the Gala' data consists of 11000 32x32 color images in 4 classes. There are 6000 training images and 5000 test images in the data.
The label classes in the dataset are:
- Food
- misc
- Attire
- Decorationandsignage
Deep Learning Problem:
- Detect and recognise the objects/images from the data set with optimum accuracy.
- Accuracy should be considerably higher.
- There is no time constraint.
Steps followed to solve the problem:
- Load and pre-process the data
- Define the model’s architecture
- Train the model
- Make predictions