Is this the intented behavior?

yeungchenwa / OCR-SAM

Combining MMOCR with Segment Anything & Stable Diffusion. Automatically detect, recognize and segment text instances, with serval downstream tasks, e.g., Text Removal and Text Inpainting

533 stars 37 forks source link

Is this the intented behavior? #26

Closed wendlerc closed 1 year ago

wendlerc commented 1 year ago

Or do I just have to configure the pipeline differently:

grafik

Mountchicken commented 1 year ago

Hi @wendlerc Sorry for the late reply. We use scene text datasets to train the text detector, thus it may not perform well on document images. Yes, this behavior is expected.