File size: 3,030 Bytes
5e4143a
68cfead
5e4143a
68cfead
 
 
de11e35
0479a69
 
 
68cfead
 
 
 
 
 
0479a69
5e4143a
68cfead
0479a69
68cfead
 
 
 
 
 
3338539
 
 
 
 
 
68cfead
 
 
 
 
 
 
3338539
 
68cfead
 
 
 
 
 
 
 
 
 
79daa12
68cfead
3338539
68cfead
3ca31ee
68cfead
 
 
 
 
 
 
 
 
 
 
 
 
 
3338539
68cfead
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
---
library_name: Doc-UFCN
license: mit
tags:
- Doc-UFCN
- PyTorch
- object-detection
- dla
- historical
- handwritten
metrics:
- IoU
- F1
- AP@.5
- AP@.75
- AP@[.5,.95]
pipeline_tag: image-segmentation
---

# Doc-UFCN - Generic historical line detection

The generic historical line detection model predicts text lines from document images.

## Model description

The model has been trained using the Doc-UFCN library on 10 historical document datasets including these public datasets:
  * [Bozen](https://zenodo.org/record/218236);
  * [cBAD2017 (READ)](https://zenodo.org/record/1491441);
  * [cBAD2019](https://zenodo.org/record/2567398);
  * [DIVA-HisDB](https://diuf.unifr.ch/main/hisdoc/diva-hisdb.html);
  * [Horae](https://github.com/oriflamms/HORAE/);
  * [ScribbleLens](https://www.openslr.org/84/).

It has been trained on images with their largest dimension equal to 768 pixels, keeping the original aspect ratio.

## Evaluation results

The model achieves the following results on the test sets:

| dataset                 | IoU   | F1    | AP@[.5] | AP@[.75] | AP@[.5,.95] |
| :---------------------- | ----: | ----: | ------: | -------: | ----------: |
| Bozen                   | 60.15 | 75.10 | 97.14   | 3.79     | 27.50       |
| cBAD2017 (READ) Complex | 46.79 | 60.35 | 56.01   | 3.40     | 16.26       |
| cBAD2017 (READ) Simple  | 53.97 | 68.43 | 57.26   | 8.45     | 19.39       |
| cBAD2019                | 50.77 | 64.52 | 35.46   | 2.88     | 11.51       |
| DIVA-HisDB              | 41.54 | 57.88 | 63.15   | 0.00     | 11.69       |
| Horae                   | 48.93 | 63.95 | 57.45   | 5.20     | 15.55       |
| ScribbleLens            | 76.61 | 86.72 | 98.02   | 71.87    | 58.32       |

The model has been trained to reduce mergers in predictions (see the [paper](https://link.springer.com/article/10.1007/s10032-022-00395-7) for more details on training). Therefore, despite slightly low evaluation values, the model correctly detects lines on a wide variety of historical and modern manuscript documents.

## How to use?

Please refer to the [Doc-UFCN library page](https://pypi.org/project/doc-ufcn/) to use this model.

## Cite us!

```bibtex
@inproceedings{boillet2022,
    author = {Boillet, Mélodie and Kermorvant, Christopher and Paquet, Thierry},
    title = {{Robust Text Line Detection in Historical Documents: Learning and Evaluation Methods}},
    booktitle = {{International Journal on Document Analysis and Recognition (IJDAR)}},
    year = {2022},
    month = Mar,
    pages = {1433-2825},
    doi = {10.1007/s10032-022-00395-7}
}
```

```bibtex
@inproceedings{doc_ufcn2021,
    author = {Boillet, Mélodie and Kermorvant, Christopher and Paquet, Thierry},
    title = {{Multiple Document Datasets Pre-training Improves Text Line Detection With
              Deep Neural Networks}},
    booktitle = {2020 25th International Conference on Pattern Recognition (ICPR)},
    year = {2021},
    month = Jan,
    pages = {2134-2141},
    doi = {10.1109/ICPR48806.2021.9412447}
}
```