Is there a way to add Image as input type to AIP logic function?
I have been looking around for a way but to no avail…
I wanted to create an AIP logic function that can generate a summarization of the content of the input image–> turns this into a function I can use with Ontology SDK-> that lets me calls a Python API and process uploaded image from user on my external application.
I read somewhere here and people said that it’s not yet possible to add Image as input type to AIP logic function. Is this true?
Media Reference inputs are not yet supported in Logic – yet! Should be coming soon. The workaround right now would be creating a Object type with a media reference property that holds the images that you would like to process. Can create this object type in Pipeline Builder (create a media set of images → convert to table rows → create object type).
You can then pass in an object / object set of that object type and then use the “Object Property” board to extract the media reference off. Afterwards, you can perform image transformations (resize image, crop image, etc) and send to a vision model to perform summarization if you’d like.