From Videos to Conversations: Egocentric Instructions for Task Assistance
Lavisha Aggarwal, Vikas Bahirwani, Andrea Colaco · 2026
Read From Videos to Conversations: Egocentric Instructions for Task Assistance by Lavisha Aggarwal, Vikas Bahirwani, Andrea Colaco (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-01038.txt.
Description
Lavisha Aggarwal and co-authors turn solo how-to clips into two-person help chats for tasks like fixes, cooking, and car work. Good task-help data is scarce, since filming real help takes time and cash, so their auto pipeline uses big models to recast one-person clips as expert-novice talks. The yield, named HowToDIV, holds 507 chats and 6,636 question-answer pairs over 24 hours of video in many fields. First scores with Gemma 3 and Qwen 2.5 give later work a mark to beat for task help in mixed media.
Catalog record
| CREATOR | Lavisha Aggarwal, Vikas Bahirwani, Andrea Colaco |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32525 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602.01038