Image caption generator

The model takes the image as an input and goes through the CNN model for feature extraction and then passes through the LSTM model for caption generation.

brown dog is running through the grass

The model will train over the training data of 30000 samples.

Insert your image here

Steps applied

1. read caption data
2. save captions into dict
3. clean text
4. create vocab
5. save descriptions
6. extract features from images using pre-trained model
7. load dataset for training the model
8. tokenize vocab into indexes
9. create data generator
10. define the model
11. Train
12. Test