tommulder commited on
Commit
94a9860
Β·
1 Parent(s): 9b8e65a

Shorter container name

Browse files
Files changed (1) hide show
  1. README.md +20 -1
README.md CHANGED
@@ -1,5 +1,5 @@
1
  ---
2
- title: "KYB YOLO-E European Document Detection"
3
  emoji: "πŸ”"
4
  colorFrom: "blue"
5
  colorTo: "purple"
@@ -19,18 +19,21 @@ This Hugging Face Space provides a production-ready API for detecting and classi
19
  ## ✨ Features
20
 
21
  ### 🎯 European Document Detection
 
22
  - **Document Types**: Identity cards, passports, driver's licenses, residence permits
23
  - **Front/Back Classification**: ML-based orientation detection using multiple methods
24
  - **Precise Coordinates**: Accurate bounding box coordinates for all detections
25
  - **Quality Assessment**: Comprehensive quality metrics (sharpness, glare, coverage, brightness, contrast)
26
 
27
  ### πŸŽ₯ Video Processing
 
28
  - **Frame Extraction**: Intelligent frame sampling at configurable FPS
29
  - **Quality-Based Selection**: Automatic selection of best quality frames
30
  - **Multi-Frame Analysis**: Track documents across video frames
31
  - **Performance Optimized**: Efficient processing for real-time applications
32
 
33
  ### πŸ”§ Technical Capabilities
 
34
  - **YOLO-E Integration**: Latest Ultralytics YOLO-E for object detection
35
  - **ML-Based Classification**: Advanced orientation detection using multiple algorithms
36
  - **European Focus**: Optimized for European document standards and formats
@@ -39,6 +42,7 @@ This Hugging Face Space provides a production-ready API for detecting and classi
39
  ## πŸš€ Quick Start
40
 
41
  ### Image Detection
 
42
  ```bash
43
  curl -X POST "https://your-hf-space-url/v1/id/detect" \
44
  -F "file=@document.jpg" \
@@ -47,6 +51,7 @@ curl -X POST "https://your-hf-space-url/v1/id/detect" \
47
  ```
48
 
49
  ### Video Detection
 
50
  ```bash
51
  curl -X POST "https://your-hf-space-url/v1/id/detect-video" \
52
  -F "file=@document_video.mp4" \
@@ -59,14 +64,17 @@ curl -X POST "https://your-hf-space-url/v1/id/detect-video" \
59
  ## πŸ“Š API Endpoints
60
 
61
  ### POST `/v1/id/detect`
 
62
  Detect European identity documents in uploaded images.
63
 
64
  **Parameters:**
 
65
  - `file` (required): Image file (JPEG, PNG, etc.)
66
  - `min_confidence` (optional): Minimum confidence threshold (0.0-1.0, default: 0.25)
67
  - `return_crops` (optional): Return cropped document images (default: false)
68
 
69
  **Response:**
 
70
  ```json
71
  {
72
  "request_id": "uuid",
@@ -102,9 +110,11 @@ Detect European identity documents in uploaded images.
102
  ```
103
 
104
  ### POST `/v1/id/detect-video`
 
105
  Detect European identity documents in uploaded videos with quality-based frame selection.
106
 
107
  **Parameters:**
 
108
  - `file` (required): Video file (MP4, AVI, etc.)
109
  - `min_confidence` (optional): Minimum confidence threshold (0.0-1.0, default: 0.25)
110
  - `sample_fps` (optional): Video sampling rate (0.1-30.0, default: 2.0)
@@ -112,6 +122,7 @@ Detect European identity documents in uploaded videos with quality-based frame s
112
  - `max_detections` (optional): Maximum detections to return (1-100, default: 10)
113
 
114
  **Response:**
 
115
  ```json
116
  {
117
  "request_id": "uuid",
@@ -125,9 +136,11 @@ Detect European identity documents in uploaded videos with quality-based frame s
125
  ```
126
 
127
  ### GET `/health`
 
128
  Health check endpoint.
129
 
130
  **Response:**
 
131
  ```json
132
  {
133
  "status": "healthy",
@@ -174,6 +187,7 @@ Each detection includes comprehensive quality assessment:
174
  ## πŸ› οΈ Configuration
175
 
176
  ### Class Mapping
 
177
  The system uses `config/labels.json` for class mapping:
178
 
179
  ```json
@@ -189,18 +203,21 @@ The system uses `config/labels.json` for class mapping:
189
  ```
190
 
191
  ### Model Weights
 
192
  - **YOLO-E**: `yolo11n.pt` (nano variant for faster inference)
193
  - **Orientation Classifier**: Integrated ML-based classification
194
 
195
  ## πŸ”§ Deployment
196
 
197
  ### Hugging Face Spaces
 
198
  1. Upload the code to a new Hugging Face Space
199
  2. Set the hardware to GPU for optimal performance
200
  3. Configure environment variables if needed
201
  4. Deploy and test the endpoints
202
 
203
  ### Local Development
 
204
  ```bash
205
  # Install dependencies
206
  pip install -r requirements.txt
@@ -212,6 +229,7 @@ python app.py
212
  ## πŸ“ Example Usage
213
 
214
  ### Python Client
 
215
  ```python
216
  import requests
217
 
@@ -231,6 +249,7 @@ for detection in result['detections']:
231
  ```
232
 
233
  ### JavaScript Client
 
234
  ```javascript
235
  const formData = new FormData();
236
  formData.append('file', fileInput.files[0]);
 
1
  ---
2
+ title: "KYB YOLO-E"
3
  emoji: "πŸ”"
4
  colorFrom: "blue"
5
  colorTo: "purple"
 
19
  ## ✨ Features
20
 
21
  ### 🎯 European Document Detection
22
+
23
  - **Document Types**: Identity cards, passports, driver's licenses, residence permits
24
  - **Front/Back Classification**: ML-based orientation detection using multiple methods
25
  - **Precise Coordinates**: Accurate bounding box coordinates for all detections
26
  - **Quality Assessment**: Comprehensive quality metrics (sharpness, glare, coverage, brightness, contrast)
27
 
28
  ### πŸŽ₯ Video Processing
29
+
30
  - **Frame Extraction**: Intelligent frame sampling at configurable FPS
31
  - **Quality-Based Selection**: Automatic selection of best quality frames
32
  - **Multi-Frame Analysis**: Track documents across video frames
33
  - **Performance Optimized**: Efficient processing for real-time applications
34
 
35
  ### πŸ”§ Technical Capabilities
36
+
37
  - **YOLO-E Integration**: Latest Ultralytics YOLO-E for object detection
38
  - **ML-Based Classification**: Advanced orientation detection using multiple algorithms
39
  - **European Focus**: Optimized for European document standards and formats
 
42
  ## πŸš€ Quick Start
43
 
44
  ### Image Detection
45
+
46
  ```bash
47
  curl -X POST "https://your-hf-space-url/v1/id/detect" \
48
  -F "file=@document.jpg" \
 
51
  ```
52
 
53
  ### Video Detection
54
+
55
  ```bash
56
  curl -X POST "https://your-hf-space-url/v1/id/detect-video" \
57
  -F "file=@document_video.mp4" \
 
64
  ## πŸ“Š API Endpoints
65
 
66
  ### POST `/v1/id/detect`
67
+
68
  Detect European identity documents in uploaded images.
69
 
70
  **Parameters:**
71
+
72
  - `file` (required): Image file (JPEG, PNG, etc.)
73
  - `min_confidence` (optional): Minimum confidence threshold (0.0-1.0, default: 0.25)
74
  - `return_crops` (optional): Return cropped document images (default: false)
75
 
76
  **Response:**
77
+
78
  ```json
79
  {
80
  "request_id": "uuid",
 
110
  ```
111
 
112
  ### POST `/v1/id/detect-video`
113
+
114
  Detect European identity documents in uploaded videos with quality-based frame selection.
115
 
116
  **Parameters:**
117
+
118
  - `file` (required): Video file (MP4, AVI, etc.)
119
  - `min_confidence` (optional): Minimum confidence threshold (0.0-1.0, default: 0.25)
120
  - `sample_fps` (optional): Video sampling rate (0.1-30.0, default: 2.0)
 
122
  - `max_detections` (optional): Maximum detections to return (1-100, default: 10)
123
 
124
  **Response:**
125
+
126
  ```json
127
  {
128
  "request_id": "uuid",
 
136
  ```
137
 
138
  ### GET `/health`
139
+
140
  Health check endpoint.
141
 
142
  **Response:**
143
+
144
  ```json
145
  {
146
  "status": "healthy",
 
187
  ## πŸ› οΈ Configuration
188
 
189
  ### Class Mapping
190
+
191
  The system uses `config/labels.json` for class mapping:
192
 
193
  ```json
 
203
  ```
204
 
205
  ### Model Weights
206
+
207
  - **YOLO-E**: `yolo11n.pt` (nano variant for faster inference)
208
  - **Orientation Classifier**: Integrated ML-based classification
209
 
210
  ## πŸ”§ Deployment
211
 
212
  ### Hugging Face Spaces
213
+
214
  1. Upload the code to a new Hugging Face Space
215
  2. Set the hardware to GPU for optimal performance
216
  3. Configure environment variables if needed
217
  4. Deploy and test the endpoints
218
 
219
  ### Local Development
220
+
221
  ```bash
222
  # Install dependencies
223
  pip install -r requirements.txt
 
229
  ## πŸ“ Example Usage
230
 
231
  ### Python Client
232
+
233
  ```python
234
  import requests
235
 
 
249
  ```
250
 
251
  ### JavaScript Client
252
+
253
  ```javascript
254
  const formData = new FormData();
255
  formData.append('file', fileInput.files[0]);