Comments (10)
Code for training will be released soon.
from manga-ocr.
Cool can you say when you have the code ready a five days or weeks?
from manga-ocr.
Hopefully within a week or so.
from manga-ocr.
Dont know if you need to label the text in the manga images to train a model, but if yes i made a darknet model to label all the manga text. Have trained it with manga109s and many other images i have labeled.
from manga-ocr.
Do you have a link to a howto that you used to make your transformers model?
from manga-ocr.
I will wait for your train code tool.
from manga-ocr.
Dont know if you need to label the text in the manga images to train a model, but if yes i made a darknet model to label all the manga text. Have trained it with manga109s and many other images i have labeled.
No need for text detection for the purpose of training OCR, but I'm thinking about going towards fully automated recognition of whole pages later, and it could be useful then.
Do you have a link to a howto that you used to make your transformers model?
I used this as a starting point:
https://github.com/NielsRogge/Transformers-Tutorials/tree/master/TrOCR
from manga-ocr.
When will you release the train code?
from manga-ocr.
I'm working on it, sorry for the delay. It's not as straightforward, because I developed it quite rapidly and now I need to clean up and document the code a bit for it to be usable at all.
from manga-ocr.
Code for training and synthetic data generation is now available.
from manga-ocr.
Related Issues (20)
- FIX: Troubleshooting for M1 MacOs users HOT 5
- caching of downloaded models HOT 2
- Failed to initialize NumPy HOT 1
- Got an error when trying to install on debian HOT 1
- TypeError: image must be numpy array type HOT 1
- ImportError: DLL load failed while importing fugashi HOT 2
- Does it work on other languages? HOT 1
- Please help, how to make it work without internet connection HOT 1
- Doesn't recognize at all HOT 1
- It downloads the model every time I run it in command line HOT 1
- Value Error HOT 1
- M1 GPU Support (MPS) HOT 2
- Output not copied to clipboard on linux HOT 2
- [offtopic] Implementation of MangaOCR in a translation software that uses GTP3.5
- error: legacy-install-failure
- error: legacy-install-failure HOT 2
- Getting terrible results on M1 mac. And I'm not sure why. HOT 1
- Is there a way to optimize transfomers backend binary size? HOT 1
- Simplified Linux clipboard support HOT 2
- Example of text which does not get recognized correctly HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from manga-ocr.