Hi all,
I'm building an app where I have to extract text from an image and I have to categorize the results.
The results that I get are quite good, but the model still makes quite a lot mistakes.
I wanted to know if the prebuilt AI models are self learning?
If not, what is the best way to improve this model.
(little remark: it's an unstructured model)
Thanx in advance!
Hi Ilias, We unfortunately do not provide options to improve the pre-built text recognition model today. Do you have examples of mistakes you're noticing in your tests? Maybe we can try to use other strategies to improve results. For example if you're looking for specific text patterns with digits only, you could apply some post processing to change Os to 0s.
Antoine