Description

Resources
 Link to the Flickr Dataset: https://www.kaggle.com/datasets/e1cd22253a9b23b073794872bf565648ddbe4f17e7fa9e
74766ad3707141adeb
 Watch this video for developing a strong intuition(pytorch implementation):
https://www.youtube.com/watch?v=y2BaTt1fxJU&t=483s Keras implementation code:
https://machinelearningmastery.com/develop-a-deep-learning-caption-generation-
model-in-python
 Follow lectures 17 to 24 in the following playlist: https://www.youtube.com/watch?v=019a30EnNGk&list=PLlH6o4fAIji5JmlmEs-
MYe1tYBwuwW5ah
 Feel free to also explore on your own and check out CHATGPT for more intuitive assistance: https://chat.openai.com/chat

Assignment : IMAGE CAPTION GENERATION

Write a code in python using either Keras or pytorch to implement VGG19 and LSTM on
FLickr8k Dataset.
Note: Make sure to use train-test ratio as 0.9:0.1 only.
Evaluation metric: BLEU-1,BLEU-2,BLEU-3,BLEU-4.