Google Vision


#1

Hi, have anyone tried the Google Vision custom activity?

I seem to get stuck on the OAuth consent screen, not sure how to fill it out.
Do you need a verified json file to use Google Vision Scope?


#2

Hello same problem here i cannot use Google Vision.

Does someone have any tip or example to show ?

Thanks !


#3

Hi @snoopy @Edupraz

See the Resources tab on Go!:



#4

Ok thank you for your help !
I’ll check this :slight_smile:


#5

Does anyone have a workflow to share?
I’ve downloaded the json but still get the message "“Could not find credentials at the path”.
I am sure it’s the right path. Maybe the json file is invalid? I did NOT fill in the consent screen because I have no clue what to write there. link

My workflow is Text Detection inside the Google Vision Scope activity. Thanks for any help on this!


#6

Hi Guys,

I make a simple hopefully fool-proof guide to get the JSON key needed for the activity.

I manage to solve the “Could not find credentials at the path problem” but now stuck at “Could not find image at the URL” problem. If anyone know how to resolve this please share it with me :slight_smile:How to activate Google Vision Activity.pdf (635.8 KB)


#7

Thanks a lot! Worked great.

In Text Detection, ImageURL:
http://wwwimages.adobe.com/www.adobe.com/content/dam/acom/en/devnet/flash/articles/page_layouts_tlf/fig02.jpg

It can’t read text in two directions though, guess we have to wait for AutoML Vision.
Demo: https://cloud.google.com/vision/


closed #8

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.