File size: 756 Bytes
c748342
 
257c25c
 
 
 
 
 
 
 
 
 
 
 
c748342
 
257c25c
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
pipeline_tag: image-segmentation
license: apache-2.0
tags:
  - vision
  - image-segmentation
  - dit
datasets:
  - ds4sd/DocLayNet-v1.1
widget:
  - src: >-
      https://upload.wikimedia.org/wikipedia/commons/c/c3/LibreOffice_Writer_6.3.png
    example_title: Wiki
---

Trained for 4 epochs.

```
model = BeitForSemanticSegmentation.from_pretrained("microsoft/dit-base", num_labels=11)
ds = load_dataset("ds4sd/DocLayNet-v1.1")
mask = np.zeros([11, 1025, 1025])
for b, c in zip(d["bboxes"], d["category_id"]):
    b = [np.clip(int(bb), 0, 1025) for bb in b]
    mask[c - 1][b[1]:b[1]+b[3], b[0]:b[0]+b[2]] = 1
mask = [cv2.resize(a, dsize=(56, 56), interpolation=cv2.INTER_AREA) for a in mask]
d["label"] = np.stack(mask)
```