Humyn Labs：为机器人构建专属互联网级训练数据
LLMs got the internet. Robots have to build their own internet.
LLMs got the internet. Robots have to build their own internet.
大语言模型有互联网可用，机器人则必须构建自己的互联网。
That is the data problem Humyn Labs is going after.
这正是Humyn Labs着手解决的数据问题。
@humynlabs is turning human experience into synchronized training data that robotics cannot readily scrape from the web.
@humynlabs 正在将人类经验转化为同步训练数据，这些数据是机器人无法轻易从网络上抓取的。
A useful robotics dataset cannot just be hours of first-person video. Humyn's samples pair human activity with signals such as IMU (inertial measurement unit), stereo depth, 6-DoF head pose, 21-point hand keypoints, wrist tracking, object tracking and dense action labels.
一个有用的机器人数据集不能仅仅是数小时的第一人称视频。 Humyn的样本将人类活动与IMU（惯性测量单元）、立体深度、6自由度头部姿态、21点手部关键点、手腕追踪、物体追踪及密集动作标签等信号配对。
Some captures even synchronize a head camera with both wrist cameras and separate IMU streams.
部分采集甚至将头戴相机与双腕相机及独立的IMU数据流同步。
So Humyn is trying to preserve enough structure around those human-demonstrations to make them useful: egocentric video, inertial measurements, hand and head pose, object trajectories, depth, narration and synchronized multi-camera views.
因此，Humyn试图在这些人类演示周围保留足够的结构信息，使其变得有用：包括自我中心视频、惯性测量、手部和头部姿态、物体轨迹、深度、叙述以及同步的多视角画面。
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力