banipreetr / Image-Captioning

Image Captioning using CNN to create embedding which is further fed to Language Generating RNN

Image-Captioning

An image to caption model which produce descriptions for real world images.

• Model architecture: CNN encoder and RNN decoder. (https://research.googleblog.com/2014/11/a-picture-is-worth-thousand-coherent.html)

Data:

Relevant links:

• train images http://msvocds.blob.core.windows.net/coco2014/train2014.zip
• validation images http://msvocds.blob.core.windows.net/coco2014/val2014.zip
• captions for both train and validation http://msvocds.blob.core.windows.net/annotations-1-0-3/captions_train-val2014.zip