Skip to content

Latest commit

 

History

History
16 lines (12 loc) · 1.24 KB

README.md

File metadata and controls

16 lines (12 loc) · 1.24 KB

A little test repo for automatic videocaptioning

Playing around with automatic image captioning using a CNN-RNN model and the Microsoft COCO dataset. Taken from https://github.com/Garima13a/Automatic-Image-Captioning

Setup

Tested with Python 3.11

  1. Install packages with pip install -r requirements.txt
  2. Get the COCO 2014 dataset from https://cocodataset.org/#home and store in a data folder (folder is defined at the top of the notebooks as datapath)
  3. Then run in following order
    1. check_coco.ipynb GoogleColab
    2. preliminaries.ipynb GoogleColab
    3. training.ipynb GoogleColab
    4. inference.ipynb GoogleColab