Exploring the power of Recurrent Neural Networks to deal with data that has temporal structure. Specifically, to generate captions for images. In order to achieve this, we will need an encoder- decoder architecture. Simply put, the encoder will take the image as input and encode it into a vector of feature values. The decoder will take this output from encoder as hidden state and starts to predict next words at each step. The following figure illustrates this:
-
Notifications
You must be signed in to change notification settings - Fork 0
jonryf/deep-learning-image-captioning
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Image captioning using vanilla RNN and Long-Term-Short-Memory (LSTM) with an encoder-decoder architecture
Resources
Stars
Watchers
Forks
Releases
No releases published