Spaces:
Sleeping
Sleeping
Update README.md
Browse files
README.md
CHANGED
@@ -1,75 +1,78 @@
|
|
1 |
-
---
|
2 |
-
title: Singtel Bill Scanner
|
3 |
-
emoji: π±
|
4 |
-
colorFrom: red
|
5 |
-
colorTo: red
|
6 |
-
sdk: streamlit
|
7 |
-
sdk_version: 1.28.0
|
8 |
-
app_file: app.py
|
9 |
-
pinned: false
|
10 |
-
tags:
|
11 |
-
- computer-vision
|
12 |
-
- ocr
|
13 |
-
- trocr
|
14 |
-
- bill-processing
|
15 |
-
- singtel
|
16 |
-
- document-ai
|
17 |
-
---
|
18 |
-
|
19 |
-
# Singtel Bill Scanner π±π‘
|
20 |
-
|
21 |
-
An AI-powered optical character recognition (OCR) system specifically designed for processing Singtel telecommunications bills.
|
22 |
-
|
23 |
-
## π Try the Live Demo
|
24 |
-
|
25 |
-
Upload your Singtel bill image and get instant results!
|
26 |
-
|
27 |
-
## Features
|
28 |
-
|
29 |
-
- π **Text Extraction**: Uses Microsoft TrOCR for accurate OCR
|
30 |
-
- π **Bill Parsing**: Extracts amounts, dates, account numbers
|
31 |
-
- β‘ **Fast Processing**: Real-time results
|
32 |
-
- π― **Singtel Optimized**: Tailored for Singtel bill formats
|
33 |
-
- π **Privacy First**: Images processed locally, not stored
|
34 |
-
|
35 |
-
## How to Use
|
36 |
-
|
37 |
-
1. πΈ Take a clear photo of your Singtel bill
|
38 |
-
2. π€ Upload the image using the interface above
|
39 |
-
3. π Click "Extract Information"
|
40 |
-
4. π View the extracted data
|
41 |
-
|
42 |
-
## Technical Details
|
43 |
-
|
44 |
-
- **Model**: Microsoft TrOCR (microsoft/trocr-base-handwritten)
|
45 |
-
- **Framework**: Streamlit + Hugging Face Transformers
|
46 |
-
- **Processing**: ~3-5 seconds per image
|
47 |
-
- **Accuracy**: High for clear, well-lit images
|
48 |
-
|
49 |
-
## Local Installation
|
50 |
-
|
51 |
-
```bash
|
52 |
-
git clone https://huggingface.co/spaces/Cosmo125/Singtel_Bill_Scanner
|
53 |
-
cd Singtel_Bill_Scanner
|
54 |
-
pip install -r requirements.txt
|
55 |
-
streamlit run app.py
|
56 |
-
```
|
57 |
-
|
58 |
-
## API Usage
|
59 |
-
|
60 |
-
```python
|
61 |
-
from transformers import pipeline
|
62 |
-
from PIL import Image
|
63 |
-
|
64 |
-
# Load model
|
65 |
-
pipe = pipeline("image-to-text", model="microsoft/trocr-base-handwritten")
|
66 |
-
|
67 |
-
# Process image
|
68 |
-
image = Image.open("bill.jpg")
|
69 |
-
result = pipe(image)
|
70 |
-
text = result[0]['generated_text']
|
71 |
-
```
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
title: Singtel Bill Scanner
|
3 |
+
emoji: π±
|
4 |
+
colorFrom: red
|
5 |
+
colorTo: red
|
6 |
+
sdk: streamlit
|
7 |
+
sdk_version: 1.28.0
|
8 |
+
app_file: app.py
|
9 |
+
pinned: false
|
10 |
+
tags:
|
11 |
+
- computer-vision
|
12 |
+
- ocr
|
13 |
+
- trocr
|
14 |
+
- bill-processing
|
15 |
+
- singtel
|
16 |
+
- document-ai
|
17 |
+
---
|
18 |
+
|
19 |
+
# Singtel Bill Scanner π±π‘
|
20 |
+
|
21 |
+
An AI-powered optical character recognition (OCR) system specifically designed for processing Singtel telecommunications bills.
|
22 |
+
|
23 |
+
## π Try the Live Demo
|
24 |
+
|
25 |
+
Upload your Singtel bill image and get instant results!
|
26 |
+
|
27 |
+
## Features
|
28 |
+
|
29 |
+
- π **Text Extraction**: Uses Microsoft TrOCR for accurate OCR
|
30 |
+
- π **Bill Parsing**: Extracts amounts, dates, account numbers
|
31 |
+
- β‘ **Fast Processing**: Real-time results
|
32 |
+
- π― **Singtel Optimized**: Tailored for Singtel bill formats
|
33 |
+
- π **Privacy First**: Images processed locally, not stored
|
34 |
+
|
35 |
+
## How to Use
|
36 |
+
|
37 |
+
1. πΈ Take a clear photo of your Singtel bill
|
38 |
+
2. π€ Upload the image using the interface above
|
39 |
+
3. π Click "Extract Information"
|
40 |
+
4. π View the extracted data
|
41 |
+
|
42 |
+
## Technical Details
|
43 |
+
|
44 |
+
- **Model**: Microsoft TrOCR (microsoft/trocr-base-handwritten)
|
45 |
+
- **Framework**: Streamlit + Hugging Face Transformers
|
46 |
+
- **Processing**: ~3-5 seconds per image
|
47 |
+
- **Accuracy**: High for clear, well-lit images
|
48 |
+
|
49 |
+
## Local Installation
|
50 |
+
|
51 |
+
```bash
|
52 |
+
git clone https://huggingface.co/spaces/Cosmo125/Singtel_Bill_Scanner
|
53 |
+
cd Singtel_Bill_Scanner
|
54 |
+
pip install -r requirements.txt
|
55 |
+
streamlit run app.py
|
56 |
+
```
|
57 |
+
|
58 |
+
## API Usage
|
59 |
+
|
60 |
+
```python
|
61 |
+
from transformers import pipeline
|
62 |
+
from PIL import Image
|
63 |
+
|
64 |
+
# Load model
|
65 |
+
pipe = pipeline("image-to-text", model="microsoft/trocr-base-handwritten")
|
66 |
+
|
67 |
+
# Process image
|
68 |
+
image = Image.open("bill.jpg")
|
69 |
+
result = pipe(image)
|
70 |
+
text = result[0]['generated_text']
|
71 |
+
```
|
72 |
+
|
73 |
+
Team members
|
74 |
+
Ninette
|
75 |
+
|
76 |
+
---
|
77 |
+
|
78 |
+
_Created for the Singtel community with β€οΈ_
|