Organization card: Studio and open datasets

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/banner.jpg filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,10 +1,51 @@
 ---
 title: README
-emoji: 🐢
-colorFrom: yellow
-colorTo: green
+emoji: 🧊
+colorFrom: gray
+colorTo: gray
 sdk: static
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+![Physicl: the worlds physical AI learns from](https://huggingface.co/spaces/physicl/README/resolve/main/assets/banner.jpg)
+
+Language models learned from the text of the internet. Physical AI, the robots and world models that have to act in real space, has nothing like that: the physical world doesn't come as a dataset. **Physicl builds it.** Real 3D environments with known geometry, materials and physics, varied as many times as a model needs, and labeled exactly by construction.
+
+## Physicl Studio
+
+Describe the situation your model gets wrong. Studio builds it from **10,000+ environments and 100,000+ assets**, varies it (light, layout, objects, clutter, camera) and renders images or video with exact ground truth on every frame.
+
+<table>
+<tr>
+<td width="50%" valign="top"><video src="https://cdn.prod.website-files.com/6abe2f7f24d2b0156b3974c2/6ae0eb95933a9a576be71cf9_blog--studio-16-variations.mp4" poster="https://cdn.prod.website-files.com/6abe2f7f24d2b0156b3974c2/6ae0eb95933a9a576be71ce5_blog--studio-16-variations-poster.jpg" autoplay loop muted playsinline width="100%"></video><br><sub>One room, sixteen versions, one camera path.</sub></td>
+<td width="50%" valign="top"><video src="https://cdn.prod.website-files.com/6abe2f7f24d2b0156b3974c2/6ae0eb96b7b9d1aaa75e0b2a_blog--studio-grid-layers.mp4" poster="https://cdn.prod.website-files.com/6abe2f7f24d2b0156b3974c2/6ae0eb95cdaceba9d8ab8b5a_blog--studio-grid-layers-poster.jpg" autoplay loop muted playsinline width="100%"></video><br><sub>The picture and its labels come out of the same render.</sub></td>
+</tr>
+</table>
+
+| On every frame | |
+|---|---|
+| Image or video | RGB render, up to 1920 × 1080 |
+| Geometry | Depth and surface normals |
+| Objects | Instance and class segmentation, 2D boxes with class and instance ID |
+| Camera | Pose for every frame |
+| Scene | Light, layout, object list and any metadata your pipeline needs |
+
+The labels come from the 3D scene itself, not estimated afterwards, so they stay exact in low light and under occlusion.
+
+**Free for research teams during the beta: [request access](https://physicl.ai/studio?utm_source=huggingface&utm_medium=org_card&utm_campaign=launch)**
+
+## Open datasets
+
+Ten rooms from the Physicl library, each rendered 300 to 400 times: 100 layouts of the same room, two framings and two lighting setups, with depth, normals, material segmentation and more on every view. Every card explains what changes, what each layer is and how to load it.
+
+<table>
+<tr><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/kitchen-workspace-understanding-safe-manipulation"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/kitchen-workspace-understanding-safe-manipulation.jpg" width="100%" alt="Kitchen Workspace Understanding & Safe Manipulation"></a><br><a href="https://huggingface.co/datasets/physicl/kitchen-workspace-understanding-safe-manipulation"><b>Kitchen Workspace Understanding & Safe Manipulation</b></a><br><sub>400 views · 11 ground-truth layers</sub></td><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/synthetic-living-room-dataset-for-robotic-perception"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/synthetic-living-room-dataset-for-robotic-perception.jpg" width="100%" alt="Synthetic Living Room Dataset for Robotic Perception"></a><br><a href="https://huggingface.co/datasets/physicl/synthetic-living-room-dataset-for-robotic-perception"><b>Synthetic Living Room Dataset for Robotic Perception</b></a><br><sub>400 views · 11 ground-truth layers</sub></td></tr>
+<tr><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/synthetic-bathroom-dataset-for-robotic-perception"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/synthetic-bathroom-dataset-for-robotic-perception.jpg" width="100%" alt="Synthetic Bathroom Dataset for Robotic Perception"></a><br><a href="https://huggingface.co/datasets/physicl/synthetic-bathroom-dataset-for-robotic-perception"><b>Synthetic Bathroom Dataset for Robotic Perception</b></a><br><sub>400 views · 11 ground-truth layers</sub></td><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/bedroom-change-detection-state-monitoring"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/bedroom-change-detection-state-monitoring.jpg" width="100%" alt="Bedroom Change Detection & State Monitoring"></a><br><a href="https://huggingface.co/datasets/physicl/bedroom-change-detection-state-monitoring"><b>Bedroom Change Detection & State Monitoring</b></a><br><sub>400 views · 11 ground-truth layers</sub></td></tr>
+<tr><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/indoor-safety-hazard-detection-and-work-zone-monitoring"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/indoor-safety-hazard-detection-and-work-zone-monitoring.jpg" width="100%" alt="Indoor Safety Hazard Detection & Work-Zone Monitoring"></a><br><a href="https://huggingface.co/datasets/physicl/indoor-safety-hazard-detection-and-work-zone-monitoring"><b>Indoor Safety Hazard Detection & Work-Zone Monitoring</b></a><br><sub>400 views · 11 ground-truth layers</sub></td><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/indoor-anomaly-detection-path-obstruction-monitoring"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/indoor-anomaly-detection-path-obstruction-monitoring.jpg" width="100%" alt="Indoor Anomaly Detection & Path Obstruction Monitoring"></a><br><a href="https://huggingface.co/datasets/physicl/indoor-anomaly-detection-path-obstruction-monitoring"><b>Indoor Anomaly Detection & Path Obstruction Monitoring</b></a><br><sub>300 views · 11 ground-truth layers</sub></td></tr>
+<tr><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/lighting-invariant-bedroom-perception-robustness-benchmark"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/lighting-invariant-bedroom-perception-robustness-benchmark.jpg" width="100%" alt="Lighting-Invariant Bedroom Perception & Robustness Benchmark"></a><br><a href="https://huggingface.co/datasets/physicl/lighting-invariant-bedroom-perception-robustness-benchmark"><b>Lighting-Invariant Bedroom Perception & Robustness Benchmark</b></a><br><sub>400 views · 11 ground-truth layers</sub></td><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/multi-view-bathroom-scene-understanding-camera-relocalization"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/multi-view-bathroom-scene-understanding-camera-relocalization.jpg" width="100%" alt="Multi-View Bathroom Scene Understanding & Camera Relocalization"></a><br><a href="https://huggingface.co/datasets/physicl/multi-view-bathroom-scene-understanding-camera-relocalization"><b>Multi-View Bathroom Scene Understanding & Camera Relocalization</b></a><br><sub>400 views · 11 ground-truth layers</sub></td></tr>
+<tr><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/indoor-scene-state-and-environmental-context-understanding"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/indoor-scene-state-and-environmental-context-understanding.jpg" width="100%" alt="Indoor Scene State & Environmental Context Understanding"></a><br><a href="https://huggingface.co/datasets/physicl/indoor-scene-state-and-environmental-context-understanding"><b>Indoor Scene State & Environmental Context Understanding</b></a><br><sub>400 views · 11 ground-truth layers</sub></td><td width="50%" valign="top"><a href="https://huggingface.co/datasets/physicl/synthetic-artifact-detection-and-domain-robustness-benchmark"><img src="https://huggingface.co/spaces/physicl/README/resolve/main/assets/synthetic-artifact-detection-and-domain-robustness-benchmark.jpg" width="100%" alt="Synthetic Artifact Detection & Domain Robustness Benchmark"></a><br><a href="https://huggingface.co/datasets/physicl/synthetic-artifact-detection-and-domain-robustness-benchmark"><b>Synthetic Artifact Detection & Domain Robustness Benchmark</b></a><br><sub>400 views · 11 ground-truth layers</sub></td></tr>
+</table>
+
+---
+
+[Website](https://physicl.ai) · [LinkedIn](https://www.linkedin.com/company/physicl) · [X](https://x.com/PhysiclAI)