-
The README says
_To help run the system, sample images from ICDAR2017 Competition on Handwritten Text Recognition on the READ Dataset are added._
I tried reproducing the results of your work by …
-
I am new in ML and learning tensorflow. I want to convert my pdf to audiobook. To remove extra text like image markings page numbers from my audio book. I want to detect the font size of this unwanted…
-
@tmbdev Hi there,
- Will it include new recognition techniques
- Will it include new text-line/paragraph segmentation methods
- Will it include export and deployment options, int8, onnx then tenso…
mrocr updated
3 years ago
-
As #5, the following documentation could be used to improve the current handwritten text recognition which is based on method that can be non-optimal for historical documents such as ours.
- [Trans…
-
Hi,
I'm trying to modify the 01-ocr notebook to work with custom dataset.
However, whenever I execute the cells with WordSegmentationNet or SegmentationNetwork the kernel dies with the following e…
-
Hi @jonomon
The number of training and testing samples comes out to be zero on running the following code in _2_line_word_segmentation.ipynb_ -
```
train_ds = IAMDataset("form_bb", output_data="bb…
-
> Please provide us with the following information:
> ---------------------------------------------------------------
### This issue is for a: (mark with an `x`)
```
- [ ] bug report -> please…
-
As the 2010’s draw to a close, it’s worth taking a look back at the monumental progress that has been made in Deep Learning in this decade.
Tags: deep_learning
via Pocket https://ift.tt/…
-
Hi, maybe I don't completely understand the structure for the required files, but I tried to guess them looking at the code.
Command I use is
`./laia/scripts/htr/train_ctc.py syms.txt [./train_ru…
-
The following documentation might be useful to implement a more precise text segmentation in order to enhance the final result :
- [A perceptive method for handwritten text segmentation](https://h…