Document Orientation for Custom Neural Extraction Models

Tyde Hashimoto 5

Hello,

I've trained a custom neural extraction model, and I've received unexpectedly poor results if the document orientation isn't correct. I had tested the documents using the prebuilt models, and they were able to pull out the data correctly; so my question is do the custom neural extraction models have a pre-processing step to fix document orientation or is that something I have to do myself?

Thank you

Pavankumar Purilla 675 Reputation points Microsoft Vendor

2024-09-27T07:07:35.9+00:00

Hi Tyde Hashimoto,

Thank you for reaching out to the Microsoft Q&A forum.

I understand that you are facing issues with your custom neural extraction model receiving unexpectedly poor results when the document orientation isn't correct.

For prebuilt models typically have additional built-in capabilities, such as handling document orientation, noise, and slight distortions.
But in custom neural extraction models doesn't have the same level of automated pre-processing. So, it's important to ensure that your documents are correctly oriented before passing them to the model.

Additionally, add more diverse data during training this helps the model generalize better.

I hope this information helps. Please do let us know if you have any further queries.
Thank you.
Pavankumar Purilla 675 Reputation points Microsoft Vendor

2024-09-30T09:39:06.13+00:00

Hi Tyde Hashimoto,
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Pavankumar Purilla 675 Reputation points Microsoft Vendor

2024-10-01T08:18:54.64+00:00

Hi Tyde Hashimoto,
Following up to see if the given response was helpful. If you need any issues, please let us know. Thank you!

Share via

Document Orientation for Custom Neural Extraction Models

Your answer