Datasets:

cybertruck32489
/

Screen2Coord

Tasks:

Image-Text-to-Text

Object Detection

Modalities:

Formats:

Sub-tasks:

visual-question-answering

instance-segmentation

Languages:

Size:

Tags:

screen-understanding

Libraries:

License:

Dataset card Data Studio Files Files and versions

cybertruck32489 commited on Sep 6

Commit

afb5f45

·

verified ·

1 Parent(s): c192f60

Upload linux-mint/dataset_info.json with huggingface_hub

Files changed (1) hide show

linux-mint/dataset_info.json +3 -9

linux-mint/dataset_info.json CHANGED Viewed

@@ -5,20 +5,14 @@
     "image": {
       "_type": "Image"
     },
-    "image_size": {
-      "feature": {
-        "dtype": "int64",
-        "_type": "Value"
-      },
-      "_type": "List"
-    },
-    "mapped_bboxes": {
       "feature": {
         "bbox": {
           "feature": {
-            "dtype": "int64",
             "_type": "Value"
           },
           "_type": "List"
         },
         "texts": {

     "image": {
       "_type": "Image"
     },
+    "mapped_denorm_bboxes": {
       "feature": {
         "bbox": {
           "feature": {
+            "dtype": "int32",
             "_type": "Value"
           },
+          "length": 4,
           "_type": "List"
         },
         "texts": {