Image captioning model, using encoder-decoder model.I used CNN as encoder and LSTM as decoder. I used VGG16 model to encode images and Glove 200-dimension vectors to encode words from captions. When given an image encoding, model predicts caption one word at a time.Evalution is done by BLEU-2 score which is 0.28 (test) and 0.39(train)
Result:
Train Result:
Test Result:
Random Result:
