Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: juliozhao/DocLayout-YOLO-DocStructBench
|
| 4 |
+
tags:
|
| 5 |
+
- object-detection
|
| 6 |
+
- document-layout-analysis
|
| 7 |
+
- onnx
|
| 8 |
+
library_name: onnx
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# DocLayout-YOLO-DocStructBench (ONNX)
|
| 12 |
+
|
| 13 |
+
[juliozhao/DocLayout-YOLO-DocStructBench](https://huggingface.co/juliozhao/DocLayout-YOLO-DocStructBench) ์ ๊ฐ์ค์น๋ฅผ **ONNX๋ก ๋ณํํ ๊ฒ**์ด๋ค.
|
| 14 |
+
๊ฐ์ค์น ์์ฒด๋ ์๋ณธ๊ณผ ๋์ผํ๋ฉฐ ์ฌํ์ตยทํ์ธํ๋ํ์ง ์์๋ค.
|
| 15 |
+
|
| 16 |
+
## ์ ONNX๋ก ๋ณํํ๋๊ฐ
|
| 17 |
+
|
| 18 |
+
์๋ณธ ๊ฐ์ค์น๋ Apache-2.0 ์ด์ง๋ง, ์ด๋ฅผ ์คํํ๋ `doclayout_yolo` ํ์ด์ฌ ํจํค์ง๋
|
| 19 |
+
**AGPL-3.0** ์ด๋ค. AGPL์ ๋คํธ์ํฌ ์๋น์ค๋ก ์ ๊ณตํ ๋๋ ์์ค ๊ณต๊ฐ๋ฅผ ์๊ตฌํ๋ ๊ฐํ
|
| 20 |
+
์นดํผ๋ ํํธ๋ผ, ์ด ํจํค์ง๋ฅผ ์ ํ์ ๋ฃ์ผ๋ฉด ๋ผ์ด์ ์ค๊ฐ ์ ์ผ๋๋ค.
|
| 21 |
+
|
| 22 |
+
Apache-2.0 ์ธ ๊ฐ์ค์น๋ง ONNX๋ก ๋ณํํด `onnxruntime`(MIT)์ผ๋ก ์คํํ๋ฉด
|
| 23 |
+
๋์ผํ ์ฑ๋ฅ์ ์ ์งํ๋ฉด์ AGPL ์ฝ๋๋ฅผ ๋ฐฐํฌ๋ฌผ์์ ์ ์ธํ ์ ์๋ค.
|
| 24 |
+
|
| 25 |
+
## ์๋ณธ ๋๋น ๋ณ๊ฒฝ์ฌํญ (Apache-2.0 ยง4 ๊ณ ์ง)
|
| 26 |
+
|
| 27 |
+
- ์๋ณธ: `juliozhao/DocLayout-YOLO-DocStructBench` (revision `8c3299a30b8ff29a1503c4431b035b93220f7b11`) ์ `doclayout_yolo_docstructbench_imgsz1024.pt`
|
| 28 |
+
- ๋ณ๊ฒฝ: PyTorch `.pt` โ ONNX ๊ทธ๋ํ๋ก ๋ด๋ณด๋ด๊ธฐ๋ง ์ํ. ๊ฐ์ค์น ๊ฐ ๋ณ๊ฒฝ ์์
|
| 29 |
+
- ๋ณํ ์ค์ : `imgsz=1024`, `opset=17`, `simplify=False`
|
| 30 |
+
- ๋ณํ ๋๊ตฌ: doclayout-yolo 0.0.4, torch 2.11.0, onnx 1.22.0
|
| 31 |
+
|
| 32 |
+
## ์
์ถ๋ ฅ
|
| 33 |
+
|
| 34 |
+
| | ํํ | ์ค๋ช
|
|
| 35 |
+
|---|---|---|
|
| 36 |
+
| ์
๋ ฅ `images` | `[1, 3, 1024, 1024]` float32 | letterbox(ํจ๋ฉ๊ฐ 114) ํ 0~1 ์ ๊ทํํ RGB |
|
| 37 |
+
| ์ถ๋ ฅ `output0` | `[1, 300, 6]` float32 | `(x1, y1, x2, y2, confidence, class_id)` |
|
| 38 |
+
|
| 39 |
+
YOLOv10 ๊ณ์ด์ด๋ผ NMS๊ฐ ๋ชจ๋ธ ์์ ์๋ค(NMS-free). ์ถ๋ ฅ์ด ์ด๋ฏธ ์ต์ข
๊ฒ์ถ ๊ฒฐ๊ณผ์ด๋ฏ๋ก
|
| 40 |
+
ํ์ฒ๋ฆฌ๋ ์ขํ ์ญ๋ณํ๊ณผ ์๊ณ๊ฐ ํํฐ๋ฟ์ด๋ค.
|
| 41 |
+
|
| 42 |
+
## ํด๋์ค (10์ข
, ์์ ๊ณ ์ )
|
| 43 |
+
|
| 44 |
+
```
|
| 45 |
+
0 title 5 table
|
| 46 |
+
1 plain text 6 table_caption
|
| 47 |
+
2 abandon 7 table_footnote
|
| 48 |
+
3 figure 8 isolate_formula
|
| 49 |
+
4 figure_caption 9 formula_caption
|
| 50 |
+
```
|
| 51 |
+
|
| 52 |
+
## ์ฌ์ฉ ์
|
| 53 |
+
|
| 54 |
+
```python
|
| 55 |
+
import cv2, numpy as np, onnxruntime as ort
|
| 56 |
+
from PIL import Image
|
| 57 |
+
|
| 58 |
+
session = ort.InferenceSession("doclayout_yolo_docstructbench_imgsz1024.onnx",
|
| 59 |
+
providers=["CPUExecutionProvider"])
|
| 60 |
+
# letterbox ์ ์ฒ๋ฆฌ๋ ultralytics ์ ๋์ผํด์ผ ์ขํ๊ฐ ๋ง๋๋ค (cv2 INTER_LINEAR, ํจ๋ฉ 114)
|
| 61 |
+
```
|
| 62 |
+
|
| 63 |
+
์ ์ฒด ๊ตฌํ์ [๊ณจ๋ ๋งํฌ OCR ๋ชจ๋](https://github.com/podongchip-stack/AIRookie/tree/feature/info/ocr)์
|
| 64 |
+
`src/goldenlink_ocr/layout.py` ์ฐธ๊ณ .
|
| 65 |
+
|
| 66 |
+
## ๋ผ์ด์ ์ค
|
| 67 |
+
|
| 68 |
+
Apache License 2.0 โ ์๋ณธ ์ ์์์ ๋ผ์ด์ ์ค๋ฅผ ๊ทธ๋๋ก ๋ฐ๋ฅธ๋ค.
|
| 69 |
+
์์ ์๊ถ์ [juliozhao/DocLayout-YOLO-DocStructBench](https://huggingface.co/juliozhao/DocLayout-YOLO-DocStructBench) ์ ์ ์์์๊ฒ ์๋ค.
|