Add Label Factory dataset 0000

0000/data/depth_estimation/dataset_info.json

@@ -1,5 +1,5 @@
 {
-  "format": "label_factory_task_parquet_v2_hf_images",
+  "format": "label_factory_task_parquet_v3_hf_images_models",
   "dataset_id": "0000",
   "task": "depth_estimation",
   "split_seed": 42,

0000/data/instance_segmentation/dataset_info.json

@@ -1,5 +1,5 @@
 {
-  "format": "label_factory_task_parquet_v2_hf_images",
+  "format": "label_factory_task_parquet_v3_hf_images_models",
   "dataset_id": "0000",
   "task": "instance_segmentation",
   "split_seed": 42,
@@ -25,5 +25,5 @@
       "test-00000-of-00001.parquet"
     ]
   },
-  "bytes": 336927007
+  "bytes": 336950277
 }

0000/data/instance_segmentation/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:003eca8eecfa8658160325cf4c378faf0efbb59b90edbe2d74fd77fa4436775b
-size 29373231
+oid sha256:0039029f4f4a382811b0e5c04967d96eb5e50b887f64675b7d00d056b85226cc
+size 29375714

0000/data/instance_segmentation/train-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6d28e712f4732bf3a84c3d3d0c00dce533b6ea2264d269dc07638284c43bccdc
-size 263824482
+oid sha256:a48ee008548195af5fb0fc6fd249b5012cca24ba219bc8dc9cd027c5a7a923d3
+size 263841934

0000/data/instance_segmentation/validation-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f704aeee9924a52ec062a6b3dae9afa2c918197a29a391373fe29074eb8e58cd
-size 43729294
+oid sha256:40a25a12c5c19719c51e57f143899d70af1be702732307fd1b1d894fb287f567
+size 43732629

0000/data/object_pose_estimation/dataset_info.json

@@ -1,5 +1,5 @@
 {
-  "format": "label_factory_task_parquet_v2_hf_images",
+  "format": "label_factory_task_parquet_v3_hf_images_models",
   "dataset_id": "0000",
   "task": "object_pose_estimation",
   "split_seed": 42,
@@ -26,5 +26,7 @@
       "test-00000-of-00001.parquet"
     ]
   },
-  "bytes": 411259902
+  "bytes": 417979754,
+  "object_models": 3,
+  "object_models_file": "models.parquet"
 }

0000/data/object_pose_estimation/models.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82ce78f97fc9f37d91706372edba15afeb0074f81a6e37ed7942b8ad22d8459e
+size 6695982

0000/data/object_pose_estimation/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2617ca85ec0af2d3dbacb1ba8d924ae6d473dac059d55dac2937cce416d9e302
-size 35901461
+oid sha256:56f634bc0dbc5da64b734e0356a7662acdc547dfd17fdc38324ecbc2944e4183
+size 35903932

0000/data/object_pose_estimation/train-00000-of-00002.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5de4ded5d2369beab0a11278d50c9dd7ced085bd137f0a11ecdbf7a0aa8f6269
-size 267031939
+oid sha256:3411fe19f0850f60018e27a01d72cb577d70c3f2782543b25eb6be4c50da499c
+size 267046823

0000/data/object_pose_estimation/train-00001-of-00002.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:40c0c15e843c84f55fd03c008e6522f28b20e2d14ea8c0b64f0541aec9e9a9f7
-size 54671158
+oid sha256:1ae4f004f3cff61b87443b85ddf2a2850db0caa5766f856da7fd0840909dbee2
+size 54674350

0000/data/object_pose_estimation/validation-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ccf1c62bfb50b3a5ab12f83bb3381eb5499b6d2ad668e1325a70fa87896e76e4
-size 53655344
+oid sha256:6b4b5e74734220d8ac2e62182b66905a882e14f8a298959f91b34a4b02d3cc27
+size 53658667

README.md

@@ -1,5 +1,5 @@
 ---
-pretty_name: Label Factory RGB-D Dataset 0000
+pretty_name: Label Factory Multi-Task RGB-D Collection
 license: other
 tags:
 - image
@@ -31,19 +31,26 @@ configs:
     path: "*/data/object_pose_estimation/validation-*.parquet"
   - split: test
     path: "*/data/object_pose_estimation/test-*.parquet"
+- config_name: object_pose_models
+  data_files:
+  - split: train
+    path: "*/data/object_pose_estimation/models.parquet"
 ---
 
-# KarolyArmin/Hand_tools: dataset 0000
+# KarolyArmin/Hand_tools
 
-Metric RGB-D scene dataset generated with the Label Factory workflow.
-The files for this capture are stored below the `0000/` directory so
-multiple numbered datasets can coexist in this repository.
+Growing multi-task RGB-D collection generated with the Label Factory workflow.
+Every capture is stored in its own zero-padded directory (`0000`, `0001`,
+`0002`, ...). Uploading a new capture appends it to the collection without
+replacing earlier directories. Task configurations aggregate compatible
+Parquet shards from every numbered directory.
 
 ## Training configurations
 
 - `depth_estimation`: RGB input, metric depth target, intrinsics and depth units
 - `instance_segmentation`: RGB input, instance-mask target, boxes and annotations
 - `object_pose_estimation`: RGB-D input, masks, camera transforms and object poses
+- `object_pose_models`: one STL payload per object type and numbered dataset
 
 Every configuration provides deterministic `train`, `validation`, and `test`
 splits. Adding another numbered capture appends rows through the wildcard paths;
@@ -112,15 +119,33 @@ camera_from_world = np.asarray(sample["camera_from_world"]).reshape(4, 4)
 world_from_objects = np.asarray(sample["world_from_objects"]).reshape(-1, 4, 4)
 camera_from_objects = np.asarray(sample["camera_from_objects"]).reshape(-1, 4, 4)
 object_names = sample["object_names"]
+
+# Resolve the STL models referenced by this pose sample.
+models = load_dataset("KarolyArmin/Hand_tools", "object_pose_models")["train"]
+required_keys = set(sample["object_model_keys"])
+models = models.filter(lambda row: row["object_model_key"] in required_keys)
+first_model = models[0]
+with open(first_model["stl_file_name"], "wb") as file:
+    file.write(first_model["stl"])
 ```
 
-## Contents
+## Collection contents
+
+Each numbered directory contributes:
+
+- RGB images and aligned 16-bit metric depth maps
+- camera intrinsics and metric world-coordinate camera poses
+- instance masks, boxes, RLE annotations and visualization overlays
+- per-frame and consolidated object poses
+- metric scene reconstruction and object STL models
+- task-specific train, validation and test Parquet shards
+
+## Latest generated contribution
 
-- 318 RGB frames and aligned metric depth maps
-- Metric camera calibration and world-coordinate camera poses
-- Metric reconstructed point cloud
-- 4 posed object instances with per-frame masks
-- STL models used for pose estimation
+- Directory: `0000/`
+- RGB-D frames: 318
+- Annotated object instances: 4
+- Selected upload files: 1945
 
 Depth images are stored in millimetres. Pose conventions and coordinate-system
 metadata are recorded in `0000/sam6d_scene/annotations.json` and