Update metadata.json

metadata.json

@@ -43,7 +43,8 @@
     "source": "cr:source",
     "subField": "cr:subField",
     "transform": "cr:transform",
-    "rai": "http://mlcommons.org/croissant/RAI/"
+    "rai": "http://mlcommons.org/croissant/RAI/",
+    "prov": "http://www.w3.org/ns/prov#"
   },
   "@type": "sc:Dataset",
   "@id": "ue5sim-renderpeople-stereo-depth-disparity",
@@ -422,5 +423,52 @@
         }
       ]
     }
+  ],
+  "rai:dataLimitations": "This dataset contains 417k stereo pairs from 128 photorealistic avatars across 109 unique stereo configurations and four indoor environments, with dense ground-truth annotations including depth\nmaps, surface normals, segmentation masks, and facial landmarks.",
+  "rai:dataBiases": "The dataset is sourced from the synthetic RenderPeople dataset of 128 individuals, male and female, over 6 ethnicities and  5 age intervals. From the perspective of 3D reconstruction and depth acquisition there shouldn't be any measurable bias.",
+  "rai:personalSensitiveInformation": "The dataset does not contain any personal sensitive information.",
+  "rai:dataUseCases": "The dataset represents 417000 stereosetups capturing human avatars placed in front of the cameras at distances varying between  30cm-300cm and dense ground-truth annotations including depth\nmaps, surface normals, segmentation masks, and facial landmarks. It is intended for evaluating methods and models at reconstructing any of : stereo setups parameters, depth maps for persons, faces and environment, surface orientation for persons, faces and environment and accurate  reconstruction of  3D objects.",
+  "rai:dataSocialImpact": "Positive impact: enables rigourous evaluation of AI or classical methods on accurate capturing of human and human faces 3D features in the context of stereoscopic setups.  Risks: none. Mitigation: the dataset is released under a non-commercial research license and requires signing a data use agreeement.",
+  "rai:hasSyntheticData": true,
+  "prov:wasDerivedFrom": [
+    {
+      "@id": "https://renderpeople.com/",
+      "prov:label": "RederPeople",
+      "sc:license": "(a) to use, reproduce, modify, create derivative works (such as body models) of, and enhance the Assets in any form or media now existing or later developed; (b) to create, use, distribute, publicly perform, license, and sub-license two-dimensional (2D) rendered images and videos containing the Assets; and (c) to create, develop, train, use, reproduce, modify, enhance, and distribute neural networks, machine learning models, and algorithms using the Assets, derivative works and (2D) rendered media thereof."
+    }
+  ],
+  "prov:wasGeneratedBy": [
+    {
+      "@type": "prov:Activity",
+      "prov:type": {
+        "@id": "https://www.wikidata.org/wiki/Q4929239"
+      },
+      "prov:label": "Data rendering",
+      "sc:description": "Extract and organise 3D models and render them in 417000 stereo setups with varying baselines, cameras characteristics, etc"
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": {
+        "@id": "https://www.wikidata.org/wiki/Q109719325"
+      },
+      "prov:label": "Automatic 3D Annotation and projection",
+      "sc:description": "annotate landmakrs on the face: eyes, lips, nose, face contour, eyebrows, etc"
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": {
+        "@id": "https://www.wikidata.org/wiki/Q1172378"
+      },
+      "prov:label": "Filter non-practical setups",
+      "sc:description": "Filter our stereo setups that are highly unlikely to be present in real cases and cases where the rendered persons does not fully fit inside the view of both cameras."
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": {
+        "@id": "https://www.wikidata.org/wiki/Q3306762"
+      },
+      "prov:label": "Final qulity review",
+      "sc:description": "manual and automatic review of the remaining stereo samples and their parameters and renders."
+    }
   ]
 }