4 ms·

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

• arnabdey0503 4mo ago
  Have you done benchmark againest other methods?
  • plexescor 4mo ago
    [dead]
  • mrkn1 4mo 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 https://huggingface.co/PaddlePaddle/PaddleOCR-VL-1.5
• explosion-s 4mo ago
  Why is no GPU a pull here? Is there usually an activation overhead GPU request or something?
  • mrkn1 4mo ago
    No, simply, my laptop only has a CPU.
• ajithhyd 4mo ago
  Can it support other languages?
  • HAWK495 4mo ago
    [flagged]
  • mrkn1 4mo ago
    It should support 109 languages. More info here: https://huggingface.co/PaddlePaddle/PaddleOCR-VL https://huggingface.co/PaddlePaddle/PaddleOCR-VL
• quanglee 4mo 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 4mo 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.