Skip to content

Dataset · lmms-lab

EgoIT-99K first-person multimodal dataset

lmms-lab/EgoIT-99K

Dataset with text, images, audio, and tables for multimodal models (100K-1M entries).

What's inside

The dataset contains multimodal data including audio, images, tables, and texts. The collection size ranges from 100 thousand to 1 million entries. Collection details and structure are described in the related EgoLife scientific paper (arXiv:2503.03803), while the language and file format in the card are not specified.

What you can do with it

  • Training multimodal models (VLM / Omni) working simultaneously with text, images, and sound
  • Research in the field of first-person perception (egocentric AI)
Size
100K–1M rows
Downloads this month
30.6K
Likes
9

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("lmms-lab/EgoIT-99K")