yutoogawa/

dataset_066192653_ocean_video_text

	dataset_066192653_ocean_video_text.py




	


	Dataset Summary

A ocean dataset with video text modality, stored in hdf5 format.

Preprocessing: domain specific Augmentation: none

Split strategy: random 90 10 Sampling: balanced

Quality filtering: strict Labeling: self training

dataset_066192653_ocean_video_text.py — main artifact of this… See the full description on the dataset page: https://huggingface.co/datasets/yutoogawa/dataset_066192653_ocean_video_text.