Comments (6)
Hello, thanks for your reply.
I have changed another way to detect (use object detection methods)
I will re-try again your method soon.
from crnn.
I retrained the model and use textgenerator, with loss 0.2.
then I try to recognize the image
the result is totally different, any suggestions?
from crnn.
- It's a bug when there aren't enough sample in the batch, there's a related issue.
- Yes, but you will have to change the
data_manager.py
file to read the labels from a file as slashes cannot be in filenames. - How long did you train, how big was your training set?
from crnn.
date test:
Recognized result: 2610m8
Recognized result: 0310tm8
I used your pre-trained weight.
I dont need the slash and I just need to number as accurate as possible, any clues?
from crnn.
If you only read dates, you can reduce the number of chars so you won't end up training letters when they can't be found in your images.
As for the slashes, if you don't want the model to predict them you can just remove them from the labels.
EDIT: Just to be clear, for the above to work you'd need to retrain the model.
from crnn.
Closing for inactivity.
from crnn.
Related Issues (20)
- How to compute accuracy both from train and test set HOT 2
- How to save model? HOT 1
- How to predict a picture HOT 1
- how many times your model iterates when you train?
- loss fall slow
- model overfitting HOT 2
- Convert code to use functional TensorFlow
- There is no run.py in the TextRecognitionDataGenerator HOT 1
- The Error rate is always 1.0 HOT 1
- fail to test
- Testing result is poor HOT 2
- CHAR_VECTOR is correct? HOT 1
- Unable to run `python3 run.py -ex ../data/test --test --restore` HOT 2
- Scaling input data between 0 and 1 HOT 2
- Default Model gives poor testing result. HOT 5
- Possible bug in using Batch normalisation HOT 1
- batch_dt = sparse_tuple_from(np.reshape(np.array(raw_batch_la), (-1)))应该去掉reshape HOT 1
- Why the output is the same when using the pretrained model for prediction?
- Problems with training Tibetan pictures HOT 3
- always training the first 100 batch
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 crnn.