# HKU's EgoWild2Dex trains dual-arm robot on 538.9 hours of "extremely dirty" human data, with under 1 hour of real-robot data

- Source: 机器之心 · 微信公众号
- Published: 2026-10-17T13:01:55.000Z
- RoboHot: https://robohot.news/en/items/iclyf8avo39lxe4qnutku894m
- Original: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2651063492&idx=3&sn=25bcc369df362a36ce14020def7e91f6&chksm=85ad4e46efc5963c69ec8dadbdab62f55706de6ff202cb17f0bee02aec0d712958546eb0c16a&scene=126&sessionid=1792243824#rd

## Summary

The University of Hong Kong's MMLab team proposes EgoWild2Dex, which directly trains a dual-arm dexterous robot on 538.9 hours of unscripted human first-person manipulation data, using under 1 hour of real-robot data per downstream task; across three long-horizon tasks—unpacking a parcel, fixing ornaments with a hot glue gun, and making a cold drink—the average success rate is 96.7%.