Hugging Face's logo

logasanjeev
/
indian-id-validator

Image-to-Text
ultralytics
PyTorch
PaddlePaddle
English
ocr
yolo
computer-vision
image-classification
object-detection
indian-id
document-processing
Eval Results (legacy)
Model card Files
xet
Community
1
logasanjeev commited on
Commit
3f082a0
·
verified ·
1 Parent(s): 728f40b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +180 -91
README.md CHANGED
@@ -15,6 +15,147 @@ tags:
15
  - YOLO
16
  - PyTorch
17
  - PaddlePaddle
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
18
  ---
19
  # Indian ID Validator
20
 
@@ -38,64 +179,18 @@ Supported ID types:
38
 
39
  ## Models
40
 
41
- ### Id_Classifier
42
- - **Model**: YOLO11l-cls
43
- - **Classes**: `aadhar_back`, `aadhar_front`, `driving_license_back`, `driving_license_front`, `pan_card_front`, `passport`, `voter_id`
44
- - **Metrics**:
45
- - Accuracy (Top-1): 0.995
46
- - Accuracy (Top-5): 1.0
47
- - **Link**: [Ultralytics Hub](https://hub.ultralytics.com/models/QnJjO78MxBaRVeX2wOO4)
48
-
49
- ### Aadhaar
50
- - **Model**: YOLO11l
51
- - **Classes**: `Aadhaar_Number`, `Aadhaar_DOB`, `Aadhaar_Gender`, `Aadhaar_Name`, `Aadhaar_Address`
52
- - **Metrics**:
53
- - mAP50: 0.795
54
- - mAP50-95: 0.553
55
- - Precision: 0.777
56
- - Recall: 0.774
57
- - Fitness: 0.577
58
- - **Link**: [Kaggle Notebook](https://www.kaggle.com/code/ravindranlogasanjeev/aadhaar)
59
-
60
- ### Driving_License
61
- - **Model**: YOLO11l
62
- - **Classes**: `Address`, `Blood Group`, `DL No`, `DOB`, `Name`, `Relation With`, `RTO`, `State`, `Vehicle Type`
63
- - **Metrics**:
64
- - mAP50: 0.690
65
- - mAP50-95: 0.524
66
- - Precision: 0.752
67
- - Recall: 0.669
68
- - **Link**: [Ultralytics Hub](https://hub.ultralytics.com/models/eaHzQ79umKwJkic9DXbm)
69
-
70
- ### Pan_Card
71
- - **Model**: YOLO11l
72
- - **Classes**: `PAN`, `Name`, `Father's Name`, `DOB`, `Pan Card`
73
- - **Metrics**:
74
- - mAP50: 0.924
75
- - mAP50-95: 0.686
76
- - Precision: 0.902
77
- - Recall: 0.901
78
- - **Link**: [Ultralytics Hub](https://hub.ultralytics.com/models/Yj4aJ34fK02MkrHFSXq0)
79
-
80
- ### Passport
81
- - **Model**: YOLO11l
82
- - **Classes**: `Address`, `Code`, `DOB`, `DOI`, `EXP`, `Gender`, `MRZ1`, `MRZ2`, `Name`, `Nationality`, `Nation`, `POI`
83
- - **Metrics**:
84
- - mAP50: 0.987
85
- - mAP50-95: 0.851
86
- - Precision: 0.972
87
- - Recall: 0.967
88
- - **Link**: [Ultralytics Hub](https://hub.ultralytics.com/models/ELaiHGZ0bbr4JwsvSZ7z)
89
-
90
- ### Voter_Id
91
- - **Model**: YOLO11l
92
- - **Classes**: `Address`, `Age`, `DOB`, `Card Voter ID 1 Back`, `Card Voter ID 2 Front`, `Card Voter ID 2 Back`, `Card Voter ID 1 Front`, `Date of Issue`, `Election`, `Father`, `Gender`, `Name`, `Point`, `Portrait`, `Symbol`, `Voter ID`
93
- - **Metrics**:
94
- - mAP50: 0.917
95
- - mAP50-95: 0.772
96
- - Precision: 0.922
97
- - Recall: 0.873
98
- - **Link**: [Ultralytics Hub](https://hub.ultralytics.com/models/jAz7y1UQAfr2oBlwLGDp)
99
 
100
  ## Installation
101
 
@@ -113,7 +208,7 @@ Supported ID types:
113
  The `requirements.txt` includes `ultralytics`, `paddleocr`, `paddlepaddle`, `numpy==1.24.4`, `pandas==2.2.2`, and others.
114
 
115
  3. **Download Models**:
116
- Models are downloaded automatically via `inference.py` from the Hugging Face repository. Ensure `config.json` is in the root directory.
117
 
118
  ## Usage
119
 
@@ -192,37 +287,27 @@ print(json.dumps(result, indent=4))
192
  **Visualizations**:
193
  The `verbose=True` flag generates visualizations for the raw image, bounding boxes, and each detected field with extracted text. Below are the results for `passport_front.jpg`:
194
 
195
- - **Raw Image**:
196
- ![Raw Image](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture1.jpeg)
197
-
198
- - **Output with Bounding Boxes**:
199
- ![Output with Bounding Boxes](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture2.jpeg)
200
-
201
- - **Detected Fields**:
202
- - **Address**:
203
- ![Address](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture9.png)
204
- - **Code**:
205
- ![Code](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture7.png)
206
- - **DOB**:
207
- ![DOB](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture4.png)
208
- - **DOI**:
209
- ![DOI](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture6.png)
210
- - **EXP**:
211
- ![EXP](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture8.png)
212
- - **Gender**:
213
- ![Gender](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture12.png)
214
- - **MRZ1**:
215
- ![MRZ1](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture13.png)
216
- - **MRZ2**:
217
- ![MRZ2](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture14.png)
218
- - **Name**:
219
- ![Name](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture10.png)
220
- - **Nationality**:
221
- ![Nationality](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture11.png)
222
- - **Nation**:
223
- ![Nation](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture3.png)
224
- - **POI**:
225
- ![POI](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture5.png)
226
 
227
  **Output**:
228
  ```
@@ -293,4 +378,8 @@ Contributions are welcome! To contribute:
293
  2. Create a branch: `git checkout -b feature-name`.
294
  3. Submit a pull request with your changes.
295
 
296
- Report issues or suggest features via the [Hugging Face Issues](https://huggingface.co/logasanjeev/indian-id-validator/discussions) page.
 
 
 
 
 
15
  - YOLO
16
  - PyTorch
17
  - PaddlePaddle
18
+ - computer-vision
19
+ - image-classification
20
+ - object-detection
21
+ - indian-id
22
+ - document-processing
23
+ model-index:
24
+ - name: Id_Classifier
25
+ results:
26
+ - task:
27
+ type: image-classification
28
+ dataset:
29
+ name: custom-indian-id-dataset
30
+ type: custom-indian-id-dataset
31
+ metrics:
32
+ - name: Accuracy (Top-1)
33
+ type: accuracy_top1
34
+ value: 0.995
35
+ - name: Accuracy (Top-5)
36
+ type: accuracy_top5
37
+ value: 1.0
38
+ source:
39
+ name: Ultralytics Hub
40
+ url: https://hub.ultralytics.com/models/QnJjO78MxBaRVeX2wOO4
41
+ - name: Aadhaar
42
+ results:
43
+ - task:
44
+ type: object-detection
45
+ dataset:
46
+ name: custom-indian-id-dataset
47
+ type: custom-indian-id-dataset
48
+ metrics:
49
+ - name: mAP50
50
+ type: mAP50
51
+ value: 0.795
52
+ - name: mAP50-95
53
+ type: mAP50-95
54
+ value: 0.553
55
+ - name: Precision
56
+ type: precision
57
+ value: 0.777
58
+ - name: Recall
59
+ type: recall
60
+ value: 0.774
61
+ - name: Fitness
62
+ type: fitness
63
+ value: 0.577
64
+ source:
65
+ name: Kaggle Notebook
66
+ url: https://www.kaggle.com/code/ravindranlogasanjeev/aadhaar
67
+ - name: Driving_License
68
+ results:
69
+ - task:
70
+ type: object-detection
71
+ dataset:
72
+ name: custom-indian-id-dataset
73
+ type: custom-indian-id-dataset
74
+ metrics:
75
+ - name: mAP50
76
+ type: mAP50
77
+ value: 0.690
78
+ - name: mAP50-95
79
+ type: mAP50-95
80
+ value: 0.524
81
+ - name: Precision
82
+ type: precision
83
+ value: 0.752
84
+ - name: Recall
85
+ type: recall
86
+ value: 0.669
87
+ source:
88
+ name: Ultralytics Hub
89
+ url: https://hub.ultralytics.com/models/eaHzQ79umKwJkic9DXbm
90
+ - name: Pan_Card
91
+ results:
92
+ - task:
93
+ type: object-detection
94
+ dataset:
95
+ name: custom-indian-id-dataset
96
+ type: custom-indian-id-dataset
97
+ metrics:
98
+ - name: mAP50
99
+ type: mAP50
100
+ value: 0.924
101
+ - name: mAP50-95
102
+ type: mAP50-95
103
+ value: 0.686
104
+ - name: Precision
105
+ type: precision
106
+ value: 0.902
107
+ - name: Recall
108
+ type: recall
109
+ value: 0.901
110
+ source:
111
+ name: Ultralytics Hub
112
+ url: https://hub.ultralytics.com/models/Yj4aJ34fK02MkrHFSXq0
113
+ - name: Passport
114
+ results:
115
+ - task:
116
+ type: object-detection
117
+ dataset:
118
+ name: custom-indian-id-dataset
119
+ type: custom-indian-id-dataset
120
+ metrics:
121
+ - name: mAP50
122
+ type: mAP50
123
+ value: 0.987
124
+ - name: mAP50-95
125
+ type: mAP50-95
126
+ value: 0.851
127
+ - name: Precision
128
+ type: precision
129
+ value: 0.972
130
+ - name: Recall
131
+ type: recall
132
+ value: 0.967
133
+ source:
134
+ name: Ultralytics Hub
135
+ url: https://hub.ultralytics.com/models/ELaiHGZ0bbr4JwsvSZ7z
136
+ - name: Voter_Id
137
+ results:
138
+ - task:
139
+ type: object-detection
140
+ dataset:
141
+ name: custom-indian-id-dataset
142
+ type: custom-indian-id-dataset
143
+ metrics:
144
+ - name: mAP50
145
+ type: mAP50
146
+ value: 0.917
147
+ - name: mAP50-95
148
+ type: mAP50-95
149
+ value: 0.772
150
+ - name: Precision
151
+ type: precision
152
+ value: 0.922
153
+ - name: Recall
154
+ type: recall
155
+ value: 0.873
156
+ source:
157
+ name: Ultralytics Hub
158
+ url: https://hub.ultralytics.com/models/jAz7y1UQAfr2oBlwLGDp
159
  ---
160
  # Indian ID Validator
161
 
 
179
 
180
  ## Models
181
 
182
+ The following models are used in the pipeline. You can download them from their respective Ultralytics Hub links in various formats such as PyTorch, ONNX, TensorRT, and more for deployment in different environments.
183
+
184
+ | Model Name | Type | Classes | Link |
185
+ |------------------|-------------------|---------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------|
186
+ | Id_Classifier | YOLO11l-cls | `aadhar_back`, `aadhar_front`, `driving_license_back`, `driving_license_front`, `pan_card_front`, `passport`, `voter_id` | [Ultralytics Hub](https://hub.ultralytics.com/models/QnJjO78MxBaRVeX2wOO4) |
187
+ | Aadhaar | YOLO11l | `Aadhaar_Number`, `Aadhaar_DOB`, `Aadhaar_Gender`, `Aadhaar_Name`, `Aadhaar_Address` | [Kaggle Notebook](https://www.kaggle.com/code/ravindranlogasanjeev/aadhaar) |
188
+ | Driving_License | YOLO11l | `Address`, `Blood Group`, `DL No`, `DOB`, `Name`, `Relation With`, `RTO`, `State`, `Vehicle Type` | [Ultralytics Hub](https://hub.ultralytics.com/models/eaHzQ79umKwJkic9DXbm) |
189
+ | Pan_Card | YOLO11l | `PAN`, `Name`, `Father's Name`, `DOB`, `Pan Card` | [Ultralytics Hub](https://hub.ultralytics.com/models/Yj4aJ34fK02MkrHFSXq0) |
190
+ | Passport | YOLO11l | `Address`, `Code`, `DOB`, `DOI`, `EXP`, `Gender`, `MRZ1`, `MRZ2`, `Name`, `Nationality`, `Nation`, `POI` | [Ultralytics Hub](https://hub.ultralytics.com/models/ELaiHGZ0bbr4JwsvSZ7z) |
191
+ | Voter_Id | YOLO11l | `Address`, `Age`, `DOB`, `Card Voter ID 1 Back`, `Card Voter ID 2 Front`, `Card Voter ID 2 Back`, `Card Voter ID 1 Front`, `Date of Issue`, `Election`, `Father`, `Gender`, `Name`, `Point`, `Portrait`, `Symbol`, `Voter ID` | [Ultralytics Hub](https://hub.ultralytics.com/models/jAz7y1UQAfr2oBlwLGDp) |
192
+
193
+ **Note**: Metrics for each model are available in the `model-index` section of the YAML metadata at the top of this README. Refer to those for detailed evaluation results.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
194
 
195
  ## Installation
196
 
 
208
  The `requirements.txt` includes `ultralytics`, `paddleocr`, `paddlepaddle`, `numpy==1.24.4`, `pandas==2.2.2`, and others.
209
 
210
  3. **Download Models**:
211
+ Models are downloaded automatically via `inference.py` from the Hugging Face repository. Ensure `config.json` is in the root directory. Alternatively, use the Ultralytics Hub links above to download models in formats like PyTorch, ONNX, etc.
212
 
213
  ## Usage
214
 
 
287
  **Visualizations**:
288
  The `verbose=True` flag generates visualizations for the raw image, bounding boxes, and each detected field with extracted text. Below are the results for `passport_front.jpg`:
289
 
290
+ | **Type** | **Image** |
291
+ |------------------------------|-----------------------------------------------------------------------------------------------|
292
+ | **Raw Image** | ![Raw Image](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture1.jpeg) |
293
+ | **Output with Bounding Boxes** | ![Output with Bounding Boxes](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture2.jpeg) |
294
+
295
+ **Detected Fields**:
296
+
297
+ | **Field** | **Image** |
298
+ |----------------|-----------------------------------------------------------------------------------------------|
299
+ | **Address** | ![Address](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture9.png) |
300
+ | **Code** | ![Code](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture7.png) |
301
+ | **DOB** | ![DOB](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture4.png) |
302
+ | **DOI** | ![DOI](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture6.png) |
303
+ | **EXP** | ![EXP](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture8.png) |
304
+ | **Gender** | ![Gender](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture12.png) |
305
+ | **MRZ1** | ![MRZ1](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture13.png) |
306
+ | **MRZ2** | ![MRZ2](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture14.png) |
307
+ | **Name** | ![Name](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture10.png) |
308
+ | **Nationality**| ![Nationality](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture11.png) |
309
+ | **Nation** | ![Nation](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture3.png) |
310
+ | **POI** | ![POI](https://huggingface.co/logasanjeev/indian-id-validator/raw/main/results/Picture5.png) |
 
 
 
 
 
 
 
 
 
 
311
 
312
  **Output**:
313
  ```
 
378
  2. Create a branch: `git checkout -b feature-name`.
379
  3. Submit a pull request with your changes.
380
 
381
+ Report issues or suggest features via the [Hugging Face Issues](https://huggingface.co/logasanjeev/indian-id-validator/discussions) page.
382
+
383
+ ## License
384
+
385
+ MIT License