Skip to content

rajesh-bhat/dhs_summit_2019_image_captioning

Repository files navigation

Image Captioning using attention models

Update: 26th July 2020

Blog published at Weights and Biases: https://app.wandb.ai/authors/image-captioning/reports/Generate-Meaningful-Captions-for-Images-with-Attention-Models--VmlldzoxNzg0ODA

Code for Image Captioning now available in TF 2.0 : tensorflow-image-captioning.ipynb

Presented at Data Hack Summit 2019 and 3rd Kaggle Days Meetup Bangalore - Senior Track

Alt text Kaggle Meetup

Task

TASK

Traing Data

TASK

Attention weight calculation

ATTENTION_WEIGHTS

Context Vector

ATTENTION_WEIGHTS

Repository referred for coding:

https://github.com/sgrvinod/a-PyTorch-Tutorial-to-Image-Captioning Modifications done: using shared weights when calculating attention weights.