We need a multi-modal data-loader that can load both time series data and recurring image data from the cloud. Moreover we need a data-loader that can work with CrossVIVIT and other TS/Image models.
[ ] Research common image/time series models and how they load data. Write this up on Confluence.
[ ] Create an ABC multi-modal loader that has methods that need to be implemented for either text, image or audio. This model should be a good base class for handling CSV or other files with images occurring at different intervals.
[ ] Create documentation detailing how files should be laid out for multi-modal loader
[ ] Create automated download of large files from GCP
We need a multi-modal data-loader that can load both time series data and recurring image data from the cloud. Moreover we need a data-loader that can work with CrossVIVIT and other TS/Image models.