Revealing the Scaling Trends of Human-Robot Co-Training. Euler Wanxiang Collaborates with Hong Kong University of Science and Technology (Guangzhou) to Open-Source Embodied UniWAM
揭示人机联合训练Scaling趋势！欧拉万象联手港科广开源具身UniWAM
Euler Wanxiang and the Hong Kong University of Science and Technology (Guangzhou) jointly released UniWAM, a Unified World and Action Model, which enables unified learning of semantic understanding, future visual prediction, and action generation. UniWAM achieved state-of-the-art results on benchmarks such as LIBERO, LIBERO-Plus, and RoboTwin 2.0, and its capabilities in instruction following, complex interaction, and long-term task execution were validated through real-world experiments. The model code and weights are fully open-sourced, supporting further research and application.
UniWAM is a unified world-action model that integrates semantic understanding, future visual prediction, and action generation, achieving state-of-the-art results on multiple benchmarks. The model's architecture enables continuous task execution and adapts to environmental changes, demonstrating the potential of human-robot co-training in embodied AI.
Source: Machine Heart — Robotics on WeChat · Read original article ↗
English translation unavailable. Choose the original article to read it.
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- code: not_reported
- weights: not_reported
- dataset: not_reported
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
提升数万小时的数据质量与数据规模 团队搭建清洗与标注管线， 处理超过一万小时机器人与人类第一视角数据，并混合 VQA 问答数据 进行预训练 。
Open source S24
UniWAM 获得了多项基准 SOTA，在 LIBERO 上取得 99.2% 的平均成功率，在 LIBERO-Plus 上取得 92.6% 的整体成功率，在 RoboTwin 2.0 C2R 上取得 68.32% 的成功率，均达到论文所列方法中的领先水平。
Open source S7
团队在所分析的数据管线中观察到：二维轨迹的形状与运动趋势往往较可靠，误差更多集中在深度和绝对尺度。
Open source S20
Source：Machine Heart — Robotics on WeChat · mp.weixin.qq.com