Datasets:

bockhealthbharath
/

Eira1-A2V-Dataset

Name: Eira1-A2V-Dataset
Creator: Bock Health

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

The notebook uploaded here is used to convert videos frames into tensor files (audio + video) which will be used by the training script to train the Video Decoder Module. The Dataset used was "https://www.kaggle.com/datasets/pevogam/ucf101". This dataset must be imported into the environment (Eg. Kaggle) and the path UCF_VIDEO_DIR must be updated in the notebook