CODEWITHSUNDEEP
azureazure-form-recognizer
roy104
roy104

Reputation: 1

Is it expected that the doc intelligence studio labelling tool only returning 2 pages of extracted value?

I try to train a custom extraction model in Azure Document Intelligence Studio on an around 10 pages document.

In their sample labeling tool, I can see the first 2 pages having the OCR extracted value for ease of labeling.

(https://i.sstatic.net/rLCAg.png)

However, from every document's 3th page onwards, no OCR extracted value is being visualized.

enter image description here

Is this an expected behaviour of the labeling tool? Such that I need to draw the labeling region myself?

Upvotes: 0

Views: 403

Answers (1)

Suresh Chikkam
Suresh Chikkam

Reputation: 3413

Is this an expected behaviour of the labeling tool? Such that I need to draw the labeling region myself?

  • There is a difference in the capabilities or limitations of the OCR extraction process. It's possible that the standard pricing tier introduced restrictions or modified settings that affected the OCR extraction beyond the second page.

Here, I have created document intelligence with standard pricing and that makes the difference in the limitation.

enter image description here

Here I have uploaded a dummy document to test the service which has 6 pages. I started analysis to check the content.

enter image description here

  • It can also detect tables, fields, figures also and we can draw manually by labelling.

enter image description here

Third pages :

enter image description here

enter image description here

Upvotes: 0

Related Questions