I am using Microsoft Document AI Intelligence to analyze some forms. I am using the prebuilt-layout model as follows.

def layout(pdf_path):
    with open(pdf_path, "rb") as f:
        poller = document_intelligence_client_sections.begin_analyze_document(
            "prebuilt-layout", f.read(), content_type="application/pdf", 
            output_content_format=ContentFormat.MARKDOWN,
    )
    result = poller.result()
    return result

However, I am getting different result for same type for forms! The forms are same but the content are different, say, Name or DoB are different. In one call, its giving sections — and within the sections its showing the paragraphs, etc. In the different call, its returning paragraphs, etc. as root level elements. I can see its a major problem, while processing forms.

Second question is, can how can I call the same with image/jpg/png instead of pdf?

Can anyone help? Thank you in advance.