Spaces:
Running
Running
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,10 +1,59 @@
|
|
| 1 |
---
|
| 2 |
title: README
|
| 3 |
-
emoji:
|
| 4 |
-
colorFrom:
|
| 5 |
colorTo: indigo
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
title: README
|
| 3 |
+
emoji: 🦾
|
| 4 |
+
colorFrom: blue
|
| 5 |
colorTo: indigo
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
+
<p align="center">
|
| 11 |
+
<b>Leverage Physical Superintelligence, with real-world data.</b>
|
| 12 |
+
</p>
|
| 13 |
+
|
| 14 |
+
<p align="center">
|
| 15 |
+
<a href="https://humaid.co">humaid.co</a> ·
|
| 16 |
+
<a href="https://www.linkedin.com/company/humaidphysicalai/">LinkedIn</a> ·
|
| 17 |
+
<a href="https://github.com/humaidtech">GitHub</a> ·
|
| 18 |
+
<a href="mailto:info@humaid.co">info@humaid.co</a>
|
| 19 |
+
</p>
|
| 20 |
+
|
| 21 |
+
---
|
| 22 |
+
|
| 23 |
+
We collect **real-world egocentric demonstration data for robotics** — across
|
| 24 |
+
manufacturing, warehouse, hospitality and food-service environments. Trained operators wear
|
| 25 |
+
calibrated multi-sensor rigs on real jobs, and we run the whole pipeline from capture through
|
| 26 |
+
annotation, quality control and delivery.
|
| 27 |
+
|
| 28 |
+
Not lab data. Not scripted. People doing their actual work.
|
| 29 |
+
|
| 30 |
+
## What we capture
|
| 31 |
+
|
| 32 |
+
| | |
|
| 33 |
+
|---|---|
|
| 34 |
+
| **Egocentric** | Stereo RGB-D, global shutter, 1920×1200 @ 30 fps |
|
| 35 |
+
| **Wrists** | Dual wrist cameras, synchronized |
|
| 36 |
+
| **Depth** | 16-bit, losslessly compressed |
|
| 37 |
+
| **IMU** | ~200 Hz, head and both wrists |
|
| 38 |
+
| **Hands** | 21-keypoint 3D model per hand |
|
| 39 |
+
| **Labels** | Temporal action segmentation + natural-language descriptions |
|
| 40 |
+
| **Format** | MCAP, every signal on a shared clock |
|
| 41 |
+
|
| 42 |
+
## Open datasets
|
| 43 |
+
|
| 44 |
+
### 🎬 [EgoViz-120](https://huggingface.co/datasets/humaidtech/EgoViz-120)
|
| 45 |
+
|
| 46 |
+
**120 hours · 5,392 clips · 7.25 TB · CC-BY-4.0**
|
| 47 |
+
|
| 48 |
+
Synchronized egocentric demonstrations of real cleaning, cooking and facility work. Stereo
|
| 49 |
+
ego video, two wrist cameras, depth, three IMUs, 3D hand pose and frame-level action labels,
|
| 50 |
+
in a single self-contained MCAP per clip.
|
| 51 |
+
|
| 52 |
+
## Working with us
|
| 53 |
+
|
| 54 |
+
We build custom datasets against a spec: environments, tasks, sensor configuration,
|
| 55 |
+
annotation schema and volume. 50+ companies served, 12+ countries, 55,000+ operators across
|
| 56 |
+
our partner networks.
|
| 57 |
+
|
| 58 |
+
If you are training manipulation or humanoid policies and the bottleneck is real-world data,
|
| 59 |
+
get in touch — [info@humaid.co](mailto:info@humaid.co).
|