Upload README.md with huggingface_hub

README.md

@@ -6,34 +6,64 @@ tags:
 - ocr
 - pytorch
 - handwritten
-license: mit
+license: apache-2.0
 datasets:
 - mrrtmob/km_en_image_line
 ---
 
 # Kiri OCR Model
 
-This is a lightweight OCR model for Kiri OCR, capable of recognizing English and Khmer text.
-Trained on the [mrrtmob/km_en_image_line](https://huggingface.co/datasets/mrrtmob/km_en_image_line) dataset.
+**Kiri OCR** is a lightweight, OCR library for **English and Khmer** documents. It provides document-level text detection, recognition, and rendering capabilities in a compact package (~13MB model).
 
-## Usage
+## ✨ Key Features
+
+- **Lightweight**: Only ~13MB model size (Lite version).
+- **Bi-lingual**: Native support for English and Khmer (and mixed).
+- **Document Processing**: Automatic text line and word detection.
+- **Robust Detection**: Works on both light and dark backgrounds (Dark Mode support).
+- **Visualizations**: Generate annotated images and HTML reports.
+
+## 📊 Dataset
+
+The model is trained on the [mrrtmob/km_en_image_line](https://huggingface.co/datasets/mrrtmob/km_en_image_line) dataset, which contains **5 million** synthetic images of Khmer and English text lines.
+
+## 💻 Usage
+
+### Installation
+
+```bash
+pip install kiri-ocr
+```
+
+### Python API
 
 ```python
-from kiri_ocr.core import OCR
+from kiri_ocr import OCR
 
-# Load from Hugging Face
-ocr = OCR(model_path="mrrtmob/kiri-ocr")
+# Initialize (loads from Hugging Face automatically)
+ocr = OCR()
 
 # Extract text
-text, results = ocr.extract_text("path/to/image.jpg")
+text, results = ocr.extract_text('document.jpg')
 print(text)
 ```
 
+### CLI Tool
+
+```bash
+kiri-ocr predict path/to/document.jpg --output results/
+```
+
 ## Model Details
-- Architecture: CRNN (CNN + LSTM + CTC)
-- Framework: PyTorch
-- Input Size: Height 32px (width variable)
+- **Architecture**: CRNN (CNN + LSTM + CTC)
+- **Framework**: PyTorch
+- **Input Size**: Height 32px (width variable)
+
+## 📈 Benchmarks
+
+Results on synthetic test images (10 popular fonts):
 
-## Benchmarks
 ![benchmark_table.png](benchmark_table.png)
+
 ![benchmark_graph.png](benchmark_graph.png)
+