-
## User Story
As an end user I have a photo that contains some text, like a signpost or billboard. I want load the photo into the app under the Vision -> OCR option, to see the photo displayed within…
-
Inference speed is mainly due to the conversion of the PaddlePaddle model to the ONNX format? Or are there other modifications made to the model itself?
-
Hello, thanks for this amazing project!
I have been using `tesseract.js` cdn in a website (hugo+webpack & netlify).
I stumbled upon your work, and found that the OCR output is significantly bette…
-
I just have tried to setup a new `ocrd_all` using Release 2023-06-14 (maybe plus some additional changes as of today).
I do a native setup on non-GPU environment (Ubuntu 22.04. using Python 3.8 via "…
-
Card to track all the job jar requests posted: https://trello.com/b/QuD7Ykrs/us-beach-job-jar
Claimed/Done:
- Set up back-end with best practices (i.e. API framework, CI pipeline, docker setup, other…
-
### Feature request
Hi, I'm one of the chinese developers that knows transformer.js by a Front End salon than invited transformer.js engineer to talk about this lib.
Now i'm going to work with clien…
-
### 在提问之前...
- [X] 我理解 Issue 是用于反馈和解决问题的,而非吐槽评论区,将尽可能提供更多信息帮助问题解决
- [X] 我填写了简短且清晰明确的标题,以便开发者在翻阅 issue 列表时能快速确定大致问题。而不是“一个建议”、“卡住了”等
- [ ] 我已查看最新测试版本的更新内容,并未提及该 bug 已被修复的情况
- [ ] 我已检查了置顶议题(pinned issu…
-
From my experience, it may be more useful to only return one result per time block of some length when searching. Currently, all frames with the text are returned, so this means there are often dozens…
-
Hi all,
From past few weeks with the help of @sayakpaul I worked on creating TFLite models for [Text Detection](https://github.com/tulasiram58827/craft_tflite) and [Recognition](https://github.com/t…
-
### What were you trying to do?
import ocrmypdf
if __name__ == '__main__': # To ensure correct behavior on Windows and macOS
ocrmypdf.ocr('线索.pdf', 'output.pdf', deskew=True)
### Where are …