breezedeus
commited on
Commit
•
d29919c
1
Parent(s):
ab085c7
Update README.md
Browse files
README.md
CHANGED
@@ -5,6 +5,7 @@ tags:
|
|
5 |
- math-formula-recognition
|
6 |
- mfr
|
7 |
- pix2text
|
|
|
8 |
- image-to-text
|
9 |
license: mit
|
10 |
library_name: transformers
|
@@ -16,11 +17,11 @@ Mathematical Formula Recognition (MFR) model from [Pix2Text (P2T)](https://githu
|
|
16 |
## Model Details / 模型细节
|
17 |
|
18 |
This MFR model utilizes the [TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr) architecture developed by Microsoft, starting with its initial values and retrained using a dataset of mathematical formula images.
|
19 |
-
The resulting MFR model can be used to convert images of mathematical formulas into LaTeX text representation.
|
20 |
|
21 |
|
22 |
此 MFR 模型使用了微软的 [TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr) 架构,以其为初始值并利用数学公式图片数据集进行了重新训练。
|
23 |
-
获得的 MFR 模型可用于把数学公式图片转换为 LaTeX
|
24 |
|
25 |
|
26 |
|
@@ -37,6 +38,7 @@ The resulting MFR model can be used to convert images of mathematical formulas i
|
|
37 |
|
38 |
## Documents / 文档
|
39 |
|
|
|
40 |
- Pix2Text (P2T) Github: [breezedeus/pix2text](https://github.com/breezedeus/Pix2Text) ;
|
41 |
- Pix2Text Online Free Service: [p2t.breezedeus.com](https://p2t.breezedeus.com/) ;
|
42 |
- Pix2Text More: [breezedeus.com/pix2text](https://breezedeus.com/pix2text) ;
|
@@ -52,7 +54,7 @@ This method doesn't need to install pix2text, but can only recognize pure formul
|
|
52 |
这种方法无需安装 pix2text,但只能识别纯公式图片。
|
53 |
|
54 |
```python
|
55 |
-
#! pip install pillow
|
56 |
from PIL import Image
|
57 |
from transformers import TrOCRProcessor
|
58 |
from optimum.onnxruntime import ORTModelForVision2Seq
|
@@ -80,11 +82,11 @@ This method requires the installation of pix2text, utilizing the Mathematical Fo
|
|
80 |
这种方法需要安装 pix2text,借助 Pix2Text 中的数学公式检测模型(MFD),它不仅可以识别纯公式图片,还可以识别包含文本的混合图片。
|
81 |
|
82 |
```bash
|
83 |
-
$ pip install pix2text
|
84 |
```
|
85 |
|
86 |
```python
|
87 |
-
#! pip install pix2text
|
88 |
|
89 |
from pix2text import Pix2Text, merge_line_texts
|
90 |
|
|
|
5 |
- math-formula-recognition
|
6 |
- mfr
|
7 |
- pix2text
|
8 |
+
- p2t
|
9 |
- image-to-text
|
10 |
license: mit
|
11 |
library_name: transformers
|
|
|
17 |
## Model Details / 模型细节
|
18 |
|
19 |
This MFR model utilizes the [TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr) architecture developed by Microsoft, starting with its initial values and retrained using a dataset of mathematical formula images.
|
20 |
+
The resulting MFR model can be used to convert images of mathematical formulas into LaTeX text representation. More detailed can be found: [Pix2Text V1.0 New Release: The Best Open-Source Formula Recognition Model | Breezedeus.com](https://www.breezedeus.com/article/p2t-v1.0).
|
21 |
|
22 |
|
23 |
此 MFR 模型使用了微软的 [TrOCR](https://huggingface.co/docs/transformers/model_doc/trocr) 架构,以其为初始值并利用数学公式图片数据集进行了重新训练。
|
24 |
+
获得的 MFR 模型可用于把数学公式图片转换为 LaTeX 文本表示。更多细节请见:[Pix2Text V1.0 新版发布:最好的开源公式识别模型 | Breezedeus.com](https://www.breezedeus.com/article/p2t-v1.0)。
|
25 |
|
26 |
|
27 |
|
|
|
38 |
|
39 |
## Documents / 文档
|
40 |
|
41 |
+
- [Pix2Text V1.0 New Release: The Best Open-Source Formula Recognition Model | Breezedeus.com](https://www.breezedeus.com/article/p2t-v1.0) ;
|
42 |
- Pix2Text (P2T) Github: [breezedeus/pix2text](https://github.com/breezedeus/Pix2Text) ;
|
43 |
- Pix2Text Online Free Service: [p2t.breezedeus.com](https://p2t.breezedeus.com/) ;
|
44 |
- Pix2Text More: [breezedeus.com/pix2text](https://breezedeus.com/pix2text) ;
|
|
|
54 |
这种方法无需安装 pix2text,但只能识别纯公式图片。
|
55 |
|
56 |
```python
|
57 |
+
#! pip install transformers>=4.37.0 pillow optimum
|
58 |
from PIL import Image
|
59 |
from transformers import TrOCRProcessor
|
60 |
from optimum.onnxruntime import ORTModelForVision2Seq
|
|
|
82 |
这种方法需要安装 pix2text,借助 Pix2Text 中的数学公式检测模型(MFD),它不仅可以识别纯公式图片,还可以识别包含文本的混合图片。
|
83 |
|
84 |
```bash
|
85 |
+
$ pip install pix2text>=1.0
|
86 |
```
|
87 |
|
88 |
```python
|
89 |
+
#! pip install pix2text>=1.0
|
90 |
|
91 |
from pix2text import Pix2Text, merge_line_texts
|
92 |
|