-
A image captioning web app using Flask
-
The web app can convert the uploaded picture into a sentence
-
The sentence will describe the image
-
ResNet50 is used as image model
-
It mainly generated the feature vectors of the images
-
LSTM is the language model
-
It is used to predict the next word from the given image and word input
-
To accomplish the purpose I have created my own vocabulary
-
The vocabulary is stored as numpy array in the vocab.npy file
-
To train the model flickr8k has been used. Link: https://www.kaggle.com/srbhshinde/flickr8k-sau
-
Keras
-
Flask
-
OpenCV
-
Colab
-
Html
-
CSS
-
Bootstrap