Hi,
I am working on a flow between a Form Processing Model in AI builder and SharePoint! When I use the prediction in the flow for my model I do not get the "request payload" fill in. I need to fill in the "document type" and "document" and I can see that it depends on the model I choose I get different fill ins. So I am wondering if you know how to fix it? guess it has something to do with my model...
Best regards,
Ina
Hello again @Anonymous,
Currently the camera control image resolution is set to 640x480. You can vote the idea proposal to increase these limits for the team to consider: https://powerusers.microsoft.com/t5/Power-Apps-Ideas/Camera-Image-Resolution/idi-p/34977
What you can use today is the Add picture control that will make use of your device camera native resolution. The following blog post explains well how to achieve this: http://www.cleverworkarounds.com/2019/06/08/getting-pics-to-sharepoint-from-powerapps-is-getting-easier/
Hope this helps!
Hi @JoeF-MSFT ,
Yes, I am using the camera control within the app to capture the image and then I have one button saying "Send to SharePoint" which is then done with a flow. The flow consists of Power App as a trigger, then the action "Convert to Binary Image" and last an action "Create File (SharePoint)"
Best regards,
Ína
Hi @Anonymous,
How are you capturing the image in the canvas app? With the camera control? And then, how are you transferring it to your SharePoint list?
Hi @JoeF-MSFT ,
There you go - I didn't get any result!
So what I am doing is that I am using a canvas app to capture an image that then is transferred to a SharePoint list so that it is possible to have an overview of all the captured pictures there. Then the pictures go to the Form Processing Model and that then identifies the key values and delivers either back to the SharePoint list or as a message to Microsoft Teams.
The pictures are within the size scope, usually between 17 - 55 KB!
However, the dimensions are not within the scope so that's probably why the error showed up! Because, when I changed the dimensions I could run the quick test and got a result. So then My question to you is, do you know how I can change the dimensions so that the image from the app is within the scope? (I read that it is possible if you have tablet layout, but I am not sure and can't find where I should change the dimensions if I have a phone layout)
Best regards,
Ína
Thanks @Anonymous.
Do you know the size of the image you are adding to the SharePoint folder that is triggering the flow?
If you take that same image from SharePoint and do a quick test, do you get results?
Hi @JoeF-MSFT,
Here you can the the flow:
And here you can see the outcome:
Hi @Anonymous,
Thanks for the update. When you go to flow edition mode can you expand the Predict action and share a screenshot? As well as copying and pasting the error message in the Predict action?
Your flow should look something like this:
Hi @JoeF-MSFT,
Thank you a lot for the answer!
I actually tried that but for some reason it doesn't work out if the trigger is "When a File is Created" in a folder (SharePoint), but as soon as I try it with the trigger "Manually Trigger a Flow".
If I use "When a File is created" as a trigger and then the "Predict" as an action the Flow Checker does not come up with any error or warning but when I test it the run fails for the "Predict" as you can see on the picture below. Test
I saw the instructions for the request payload here: https://www.youtube.com/watch?v=2vnKw-w9EUM
Hi @Anonymous,
Thanks for posting the question.
Since early February 2020 there is no request payload to fill in the predict action, instead you now specify the Document type (image/jpeg for jpeg images, image/png for png images or application/pdf for pdf files) and the content of the form to analyze on the Document field.
You can find detailed documentation on how to use your form processing models in Power Automate here: https://docs.microsoft.com/en-us/ai-builder/form-processing-model-in-flow
You can also quickly get started using the following Power Automate template: https://us.flow.microsoft.com/en-us/galleries/public/templates/da60c101494a4083b292e327f2443/click-a-button-to-read-and-save-information-from-documents-using-ai-builder/
For my knowledge, can you let me know where you have seen instructions referring to request payload?
Let us know if this helps or we can further assist!
WarrenBelz
146,635
Most Valuable Professional
RandyHayes
76,287
Super User 2024 Season 1
Pstork1
65,997
Most Valuable Professional