MobileNet CNN model trained as a Non-essential whatsapp images classifier
- MobileNet CNN model trained on a custom dataset of about 5000 whatsapp images.
- The model classifies the images into 6 categories namely Paper, Screenshots, Family, Posters, Memes and Others.
- The idea is to categorize the images and select non-essential whatsapp images among these categories based on confidence scores which can then be deleted in one go instead of having to select them manually for deletion.
The model achieved a final accuracy of 94.66% after performing transfer learning on a COCO dataset trained MobileNet model.