Commit
•
df2c601
1
Parent(s):
b32907b
Upload folder using huggingface_hub (#1)
Browse files- Upload folder using huggingface_hub (a7a9b71009423ceb7bc9860300fc0be521de2eca)
- Update README.md (bb2c0c1d93f899bad410b3b00972083377126f72)
Co-authored-by: Yoann Schneider <[email protected]>
README.md
CHANGED
|
@@ -1,3 +1,50 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: YOLO
|
| 3 |
+
license: cc0-1.0
|
| 4 |
+
tags:
|
| 5 |
+
- YOLO
|
| 6 |
+
- object-detection
|
| 7 |
+
- dla
|
| 8 |
+
- historical
|
| 9 |
+
metrics:
|
| 10 |
+
- P
|
| 11 |
+
- R
|
| 12 |
+
- mAP@50
|
| 13 |
+
- mAP@50-95
|
| 14 |
+
pipeline_tag: object-detection
|
| 15 |
+
version:
|
| 16 |
+
- YOLOv8
|
| 17 |
+
---
|
| 18 |
+
|
| 19 |
+
# YOLOv8 - PELLET Casimir Marius - Line + Illustration detection
|
| 20 |
+
|
| 21 |
+
This model predicts the following elements from document images:
|
| 22 |
+
- text line
|
| 23 |
+
- illustration
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
The model has been trained using the YOLOv8 library on [PELLET Casimir Marius](https://huggingface.co/datasets/Teklia/PELLET-Casimir-Marius-yolov8) dataset.
|
| 28 |
+
It has been trained on images with their dimensions equal to 640 pixels, starting from the YOLOv8m checkpoint.
|
| 29 |
+
|
| 30 |
+
## Evaluation results
|
| 31 |
+
|
| 32 |
+
The model achieves the following results:
|
| 33 |
+
|
| 34 |
+
| Set | Class | Images | Instances | P | R | mAP@50 | mAP@50-95 |
|
| 35 |
+
| ----- | ------------ | ------ | --------- | ----- | ----- | ------ | --------- |
|
| 36 |
+
| train | all | 80 | 881 | 0.707 | 0.823 | 0.825 | 0.457 |
|
| 37 |
+
| | text line | 80 | 843 | 0.801 | 0.73 | 0.781 | 0.275 |
|
| 38 |
+
| | illustration | 80 | 38 | 0.613 | 0.916 | 0.87 | 0.64 |
|
| 39 |
+
| val | all | 10 | 129 | 0.608 | 0.872 | 0.847 | 0.493 |
|
| 40 |
+
| | text line | 10 | 125 | 0.791 | 0.744 | 0.7 | 0.256 |
|
| 41 |
+
| | illustration | 10 | 4 | 0.424 | 1 | 0.995 | 0.731 |
|
| 42 |
+
| test | all | 10 | 125 | 0.767 | 0.881 | 0.932 | 0.556 |
|
| 43 |
+
| | text line | 10 | 122 | 0.903 | 0.763 | 0.868 | 0.331 |
|
| 44 |
+
| | illustration | 10 | 3 | 0.631 | 1 | 0.995 | 0.78 |
|
| 45 |
+
|
| 46 |
+
## How to use?
|
| 47 |
+
|
| 48 |
+
- Download the [weights of this model](https://huggingface.co/Teklia/PELLET-CasimirMarius-YOLOv8/resolve/main/model.pt?download=true)
|
| 49 |
+
- Refer to the [Ultralytics documentation](https://docs.ultralytics.com/modes/predict/) to use this model.
|
| 50 |
+
|
model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c25d70ce998283ecbaf6a10fe928fcc871c044cf35f975a94ce8abcc417d63f
|
| 3 |
+
size 54844747
|