Show HN: CPU-only fast OCR for screenshots, images, PDFs, webpages

Show HN: CPU-only fast OCR for screenshots, images, PDFs, webpages(github.com)

9 points by mrkn1 94 days ago | 8 comments

quanglee 92 days ago |

Do you have some benchmark on how fast it is on cpu? I checked the repo and the huggingface page but only found gpu-based benchmarks

mrkn1 91 days ago | |

I have not. My experience has been a few seconds for an 1024x1024 with medium density of text, FWIW. Feel free to try it on a few test images, model is pretty small and fast, but yeah no formal evals on CPU.

explosion-s 94 days ago |

Why is no GPU a pull here? Is there usually an activation overhead GPU request or something?

mrkn1 94 days ago | |

No, simply, my laptop only has a CPU.

arnabdey0503 94 days ago |

Have you done benchmark againest other methods?

mrkn1 94 days ago | |

I haven't. But the evals of the underlying model are published here, including on Omnibench. https://huggingface.co/PaddlePaddle/PaddleOCR-VL-1.5

ajithhyd 93 days ago |

Can it support other languages?

mrkn1 93 days ago | |

It should support 109 languages. More info here: https://huggingface.co/PaddlePaddle/PaddleOCR-VL