Resolving Schema Inconsistency Issues

Using Labelbox Annotate
1

Firstly, thanks for the awesome Labelbox. Since the error doesn’t seem to be serious, I decided to share it here in the community.

I’ve already exported three batches of annotation projects. However, just to obtain fresh annotations from my dataset, I ran a new export, this time with version 2. While it was seamless on the latest one, I encountered an error in my oldest annotation project:

{"error": "SchemaInconsistencyException", "message": "Unrecognized schema_id: cldg4y8rn249u07z54sya4ctr from dict_keys(['cl7gti0ad0n0u072r25my7zls', 'cl7gti0ad0n0w072r9ve2ai00', 'cl7gti0ad0n0y072r2mv44enp', 'cl7gti0ad0n10072r5c940ot4', 'cl7gti0ad0n12072r9wwqeenm', 'cl7gti0ad0n14072ralup6kbi', 'cl7gti0ad0n0v072r0b0s8x0f', 'cldgcf9f9012m07zg98m08o7f', 'cldgcf9f9012o07zg6nry1cfl', 'cldgcf9f9012q07zg50p610zj', 'cldgcf9f9012s07zgdr901w2y', 'cldgcf9fa012u07zg4ch211ds', 'cldgcf9fa012w07zgf8n106e0', 'cldgcf9f9012n07zgdxkk7nkf', 'ckrdp1ln400003h6jq23cbjfl'])", "reason and fix": "the feature/ tool mentioned in the error message for the label cldg4frua25do07vdhmc651c6 is not linked to the Ontology. Attach the feature to the ontology and try exporting again. Alternatively, you could get rid of the annotation by doing this: 1. attach the feature to the Ontology, 2. go to the label editor, delete the annotation from the datarow,3. detach the feature from the Ontology and export again.", "label_id": "cldg4frua25do07vdhmc651c6", "data_row_id": "cl7fjb9do038l075rfclv5syw", "task_id": null}
{"error": "SchemaInconsistencyException", "message": "Unrecognized schema_id: cldg4y8rn249u07z54sya4ctr from dict_keys(['cl7gti0ad0n0u072r25my7zls', 'cl7gti0ad0n0w072r9ve2ai00', 'cl7gti0ad0n0y072r2mv44enp', 'cl7gti0ad0n10072r5c940ot4', 'cl7gti0ad0n12072r9wwqeenm', 'cl7gti0ad0n14072ralup6kbi', 'cl7gti0ad0n0v072r0b0s8x0f', 'cldgcf9f9012m07zg98m08o7f', 'cldgcf9f9012o07zg6nry1cfl', 'cldgcf9f9012q07zg50p610zj', 'cldgcf9f9012s07zgdr901w2y', 'cldgcf9fa012u07zg4ch211ds', 'cldgcf9fa012w07zgf8n106e0', 'cldgcf9f9012n07zgdxkk7nkf', 'ckrdp1ln400003h6jq23cbjfl'])", "reason and fix": "the feature/ tool mentioned in the error message for the label cldg4frua25do07vdhmc651c6 is not linked to the Ontology. Attach the feature to the ontology and try exporting again. Alternatively, you could get rid of the annotation by doing this: 1. attach the feature to the Ontology, 2. go to the label editor, delete the annotation from the datarow,3. detach the feature from the Ontology and export again.", "label_id": "cldg4frua25do07vdhmc651c6", "data_row_id": "cl7fjb9do038l075rfclv5syw", "task_id": null}

This is similar to this thread asked before. As mentioned, I already have my batches of annotations, and they seem fine after some cleaning.

I guess it may be due to the migrated catalog or some problem with the ontology. My problem is that I can not pinpoint exactly which video/annotation/feature is problematic or how to address it.

2

Hey @Didas ,

From the error message :

 "label_id": "cldg4frua25do07vdhmc651c6",
 "data_row_id": "cl7fjb9do038l075rfclv5syw"

This is your label / data row id where the issue lies, to solve this you would need to add the feature specified to have a “clean” export.

The Feature (tool) mention in the error is :

{"title":"seg","value":"seg","color":"#FF4A46"}

It’s not deleted so it should be straighforward.

  1. Edit the existing ontology and add this feature
  2. Either delete the annotations (of not need it) or keep as is
  3. Export without issue :crossed_fingers:

Let me know how it goes!

Many thanks,
PT

1 Like