Organize the data by task

README.md

@@ -11,6 +11,7 @@ task_categories:
 - any-to-any
 configs:
 - config_name: birds_eye_view_rotation
+  default: true
   data_files:
   - split: train
     path: preview/birds_eye_view_rotation.parquet
@@ -26,7 +27,7 @@ configs:
   data_files:
   - split: train
     path: preview/jigsaw_single_image.parquet
-- config_name: bbox
+- config_name: visual_search_bbox
   data_files:
   - split: train
     path: preview/bbox.parquet
@@ -111,7 +112,8 @@ the paper: basic visual perception, static spatial understanding, and mental wor
 |  | `exo_to_ego` | the person's own view, from an outside view | — | 4,526 |
 | **Total** | | | **142,333** | **38,157** |
 
-All rows are training rows. Choose a task in the Dataset Viewer to see examples.
+All rows are training rows. Choose a task in the Dataset Viewer to see examples; there `bbox` is
+listed as `visual_search_bbox`.
 
 ## Files
 
@@ -195,7 +197,26 @@ To obtain the source data:
   image record names the take and the camera. The frame-selection code will be released with
   the training code.
 
-## License
-
-The images come from several public datasets, and each keeps its own terms. There is no blanket
-license for the images. A list of the image sources per task will be added with the paper.
+## Sources and licenses
+
+Images keep the terms of the dataset they come from; check them before you use the data.
+Several sources allow non-commercial research use only. Licenses below are as stated by each
+source.
+
+| Task | Image sources | Terms |
+| --- | --- | --- |
+| `bbox` | [SA-1B](https://ai.meta.com/datasets/segment-anything/); [Visual-CoT](https://huggingface.co/datasets/deepcs233/Visual-CoT) (its images come from COCO, GQA, TextVQA, Visual Genome, Flickr30k, DocVQA, Open Images, CUB and others) | SA-1B Dataset Research License; Visual-CoT: CC BY-NC 4.0, images under their original terms |
+| `crop`, `jigsaw_single_image` | [SA-1B](https://ai.meta.com/datasets/segment-anything/) | SA-1B Dataset Research License |
+| `jigsaw_multi_frame` | [Hypersim](https://github.com/apple-aiml-research/ml-hypersim), [ScanNet](https://github.com/ScanNet/ScanNet), [Open X-Embodiment](https://github.com/google-deepmind/open_x_embodiment) robot datasets, [Visual7W](https://github.com/yukezhu/visual7w-toolkit) (COCO images) and other open images | Hypersim: CC BY-SA 3.0; ScanNet: ScanNet Terms of Use; Open X-Embodiment: CC BY 4.0; COCO images: Flickr terms |
+| `object_segmentation` | [LVIS v1](https://www.lvisdataset.org/) on [COCO 2017](https://cocodataset.org/#termsofuse) images | LVIS annotations: CC BY 4.0; COCO images: Flickr terms |
+| `depth_map`, `marked_depth_map` | [RefSpatial](https://huggingface.co/datasets/JingkunAn/RefSpatial): [Open Images](https://storage.googleapis.com/openimages/web/index.html), [CA-1M](https://github.com/apple/ml-cubifyanything) and simulated scenes | RefSpatial: Apache-2.0; Open Images: CC BY 2.0; CA-1M: CC BY-NC-ND |
+| `object_rotation` | [MVImgNet](https://github.com/GAP-LAB-CUHK-SZ/MVImgNet) | MVImgNet Terms of Use (non-commercial research) |
+| `object_editing` | synthetic scenes in the style of [CLEVR](https://cs.stanford.edu/people/jcjohns/clevr/) | CLEVR: CC BY 4.0 |
+| `birds_eye_view`, `birds_eye_view_rotation` | [MindCube](https://huggingface.co/datasets/MLL-Lab/MindCube) (its images come from [ARKitScenes](https://github.com/apple-aiml-research/ARKitScenes), WildRGB-D, DL3DV-10K and self-collected scenes) | MindCube: MIT; ARKitScenes: Apple license, non-commercial |
+| `panoramic_completion` | [Realsee3D](https://doi.org/10.5281/zenodo.17826243), [360+x](https://x360dataset.github.io/); [Matterport3D](https://niessner.github.io/Matterport/) (text only) | Realsee3D: CC BY 4.0; 360+x: CC BY-NC-SA 4.0; Matterport3D: Terms of Use |
+| `object_viewpoint_render`, `object_viewpoint_render_oriented`, `object_orientation` | [COCO 2017](https://cocodataset.org/#termsofuse) images | Flickr terms (COCO terms of use) |
+| `marked_position_view` | [SAT](https://huggingface.co/datasets/array/SAT) (rendered rooms) | MIT |
+| `exo_to_ego` | [Ego-Exo4D](https://ego-exo4d-data.org/) (text only) | Ego-Exo4D license agreement |
+
+The reasoning text, the intermediate images and the release layout are ours. There is no
+blanket license over the images.

preview/jigsaw_multi_frame.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c771a748b5c25cb79382b0d12314b46727fe8c02042fff12e80a0089dab42277
-size 1442880
+oid sha256:1fdc3d10b2140454538689aad4259a50fad76ba1b2a17e53b2acaeaacb6e12c3
+size 692052

shards/SHARDS.json

@@ -1,14 +1,14 @@
 {
   "schema_version": 1,
-  "subset": "primer-instruct-181k",
+  "subset": "primer-instruct-180k",
   "release_ready": false,
-  "manifest_sha256": "8dd89adca1ab68e0362a4e3650dfc4910a6f058213c00bfacb456dd12cbe6859",
+  "manifest_sha256": "80923b55de5dddc81e5b99b7f5b7784f38c96be52e6554d47a7a8f5b8b18dbdf",
   "shard_target_bytes": 1073741824,
   "files": [
     {
       "path": "annotations.tar",
       "bytes": 464742400,
-      "sha256": "89eb99a41a70f889edf44ee733d7292cdbb7f810fbc97e49550e68791e27cbe4"
+      "sha256": "3eea5452b29d19c74a4c78abd07b5356ec12b19fe9ee3325bb051669af928939"
     },
     {
       "path": "data-00000.tar",

shards/annotations.tar

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:89eb99a41a70f889edf44ee733d7292cdbb7f810fbc97e49550e68791e27cbe4
+oid sha256:3eea5452b29d19c74a4c78abd07b5356ec12b19fe9ee3325bb051669af928939
 size 464742400

text_only/FILES.json

@@ -1,6 +1,6 @@
 {
   "schema_version": 1,
-  "subset": "primer-instruct-181k/text_only",
+  "subset": "primer-instruct-180k/text_only",
   "files": [
     {
       "path": "exo_to_ego/train.jsonl",
@@ -25,7 +25,7 @@
     {
       "path": "MANIFEST.json",
       "bytes": 1385,
-      "sha256": "41f334ebd4b6a8e428ba20aa67fc1f64e376bcb4a492e55584ac240e2f4afb5e"
+      "sha256": "b44a193e45147c1cf57b0de87feb574373c6d1c61df40ff0bb6ff2eab131249f"
     }
   ]
 }

text_only/MANIFEST.json

@@ -1,6 +1,6 @@
 {
   "schema_version": 1,
-  "corpus": "primer-instruct-181k",
+  "corpus": "primer-instruct-180k",
   "package": "text_only",
   "layout": "tasks",
   "rows": {