HTRPipeline
HTRPipeline copied to clipboard
Failing while using test data
I'm a beginner in ML field, I followed the steps and successfully ran the model, but it fails on simple test data. Could I be missing something?
it is due to the word detector part, the model is not trained to detect such big words (most likely the image has a high resolution). try either the scale parameter (first slider in the web UI), or try manually downscaling the image.
Does the model recognize symbols?
in the file reader.json (which you get when you download the model) you see the list of all characters that the model can read