Brian Davis

Would anyone mind reporting your best ANLS scores using Tesseract? @dongxuewang-123 their server had a bug which should be fixed now

In case this helps someone, I adapted the mentioned collab into a training script for the full dataset. It can train either on Tesseract OCR or the dataset OCR. It...

Is this the pretrained detector, or your own?

Whoops. The detector snapshot is actual another snapshot of the pairing model. I'll fix that.

Ok, re-download the weights. It has the correct snapshot now.

If running the detection-only model, `detect_boxes()` returns the bounds boxes, you just need to add code to save them to a file in the format you desire. If running pairing...

Sorry I never saw this! You can use the new `run.py` script with the `-d` flag to run detection on a single image now.

Yes, training is slow. It took me a couple days for the detector. The pairing network takes even longer, on the order of a couple weeks. I'll also point you...

I'm not sure I understand what's going on. It looks like the model was run on 50 images (wrote 50 different jsons). What to you mean by "I was able...

What is the specific command your using with `eval.py`? And what is your desired output? Is the jsons what you want, or is it something else?