Comments (8)
Yes, in some languages, the results we obtained were below the baseline, but our method is very fast, easy to train and language-independent. Also, notice from Table 1 that we obtained this results with as few as 200 documents.
I don't see how you can build a weighting function using word frequency: remember that the weighting function is what determines the weight of each character in each n-gram when you join them together.
from post_ocr_correction.
The idea of using word frequency to calculate the weight function came from OCR error correction using correction patterns and self‑organizing migrating algorithm. Besides, I hava a question about the challenge of Post-OCR Text Correction. Why doesn't the authorities provide the image data? I think image features are crucial for OCR text correction tasks. Is there any way to get the image data?
from post_ocr_correction.
Challenge ICDAR2019 Competition on Post-OCR Text Correction
from post_ocr_correction.
The whole purpose of the competition is to correct text when the documents have already been scanned and the image of the document itself has been lost. There is no way to obtain the image data for that particular competition.
from post_ocr_correction.
Got it
from post_ocr_correction.
Got it
Hi Qingkun, do you know how to do all the training all over again? Did you try it?
The author said we can go to notebooks, but the code is outdate and we need to change many things. Have you already change the code in notebooks? Thanks!
from post_ocr_correction.
Got it
Hi Qingkun, do you know how to do all the training all over again? Did you try it?
The author said we can go to notebooks, but the code is outdate and we need to change many things. Have you already change the code in notebooks? Thanks!
Try this https://github.com/xuqinkun/post_ocr_correction
from post_ocr_correction.
Got it
Hi Qingkun, do you know how to do all the training all over again? Did you try it?
The author said we can go to notebooks, but the code is outdate and we need to change many things. Have you already change the code in notebooks? Thanks!
Qingkun, I still cannot run it. Could you please try to run it with latest python version and give an example command? Thanks!
This is the command I use, is it correct?
python ./train/train_from_scratch.py -i ./train/data/ICDAR2019_POCR_compttition_dataset -ds SROIE -m ./train/data/models/en/model_en.pt -o ./ --do_train --overwrite_model --overwrite_eval --overwrite_train
from post_ocr_correction.
Related Issues (9)
- How to train with other language ? HOT 6
- Will you able to share the pretrained model? HOT 1
- vocabulary not found HOT 1
- Dealing with long text HOT 2
- ModuleNotFoundError: No module named 'pytorch_decoding' HOT 1
- evaluate_model HOT 3
- How can I train the model with the dataset in the paper? HOT 4
- For the inference, how can we do the alignment? HOT 2
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 post_ocr_correction.