3 ms·
I'm working on a projet that uses PaddleOCR to get bounding boxes. It's far from perfect, but it's open source and good enough for our requirements. And it can
by yfontana 2y ago
I'm working on a projet that uses PaddleOCR to get bounding boxes. It's far from perfect, but it's open source and good enough for our requirements. And it can mostly handle a 150 MB single-page PDF (don't ask) without completely keeling over.