Skip to content

We compare 2 widely used image captioning methods, with and without attention

Notifications You must be signed in to change notification settings

Nilanjan96/Comparison-of-image-captioning-methods

Repository files navigation

Comparison-of-image-captioning-methods

We compare 2 widely used image captioning methods, with and without attention

ImageCaptioningBasicBest.ipynb:- This contains the best model we got so far by training on flickr8k dataset (70% train and 30% test sets). For generation, we have used greedy search here.

Image_Captioning.ipynb:- This notebook contains implementation of image captioning with visual attention

ImageCaptioningBasicGreedy_Beam.ipynb:- This notebook includes our implementation of beam search and compares the performance of greedy and beam search in generation of sentence.

ImageCaptioning.ipynb:- Here, we have implemented the model and trained on just 100 images from the flickr30k dataset. This was to test the working soundness of the architecture

About

We compare 2 widely used image captioning methods, with and without attention

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published