Tencent Robot X Team Creates 'Thinking and Imagining' Robot Brain - RxBrain
腾讯机器人X团队打造“会思考也会想象”的机器人大脑——RxBrain
The Tencent Robot X Team, Fujita Lab, and Tencent Hunyuan Team jointly released a study proposing a new embodied cognition foundational model called Hy-Embodied-RxBrain (shortened as RxBrain). This model aims to enable robots to not only describe steps in language when planning actions but also generate mental images of the world state after each step in their mind.
Source: Chinese robotics — Dataset and collection discovery · 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.
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
原始视频数据来自四大类来源：真实机器人操作数据（约17292小时）、UMI手持夹持器采集数据（约17506小时）、仿真环境数据（约1317小时）、以及第一人称人类日常操作视频（约14062小时）。
Open source S18
总计超过5万小时，包含约2150万个可用片段，通过约2861万个候选片段筛选而来，通过率约75%。
Open source S19
最终，约75%的候选片段通过了质量验证。
Open source S20
第一阶段是联合视觉语言预训练，训练数据由两大部分混合而成：60%是大规模图文对（来自LAION和COYO数据集，共约2亿条），40%是多模态指令跟随数据（来自LLaVA-OneVision和MAmmoTH，共约1.13亿条）。
Open source S22
第二阶段是具身监督微调，从第一阶段的检查点继续训练，训练数据混合了六类内容，总计约422份的混合比例：一般图文对（T2I基础数据和高质量图文对，合计约25.9%）、通用多模态理解数据（34.6%）、具身多帧世界模型数据（16.2%）、具身交错规划数据（20.8%）。
Open source S23
Source：Chinese robotics — Dataset and collection discovery · sohu.com