Hey there Vikas M
thats a great question thanks for using QandA platform
As you have said you already trained the models, you can test it with sample documents to check if it accurately detects signatures in the specified fields. Make sure to provide diverse samples covering different types of signatures and variations.
Review the results generated by the model to determine if signatures are correctly identified in the designated fields. The model's accuracy may vary based on factors such as the quality of training data, complexity of signatures, and document variations.
If the model does not perform as expected**, you may need to iterate on the training process, refine the labeling of signature fields,** or provide additional training data to improve accuracy.
Speakimg of alternatives, If Azure AI Document Intelligence does not meet your requirements even after following everything , you can explore other alternatives such as custom machine learning models trained specifically for signature detection.
I can assure you to get accurate results you have to do a lot of refinements, all the very best
If this helps kindly accept the answer thanks much.