Upload README.md with huggingface_hub

README.md

@@ -12,11 +12,11 @@ tags:
 
 # Qalam-Net (قلم-نت): Advanced Arabic OCR (v2 Portable)
 
-Qalam-Net is a high-performance, cross-backend Optical Character Recognition (OCR) model for Arabic. Built on **Keras 3**, it supports **JAX**, **PyTorch**, and **TensorFlow** backends.
+Qalam-Net is a high-performance, cross-backend Optical Character Recognition (OCR) model for Arabic. Patched for **Keras 3**, it supports **JAX**, **PyTorch**, and **TensorFlow**.
 
 ## 🚀 Quick Start (Robust Usage)
 
-For maximum reliability across all environments, use the `huggingface_hub` downloader.
+This guide uses a custom **NumPy-based decoder** to ensure compatibility across all Keras 3 backends without needing `tf.keras.backend.ctc_decode`.
 
 ### 1. Installation
 ```bash
@@ -35,37 +35,43 @@ from huggingface_hub import hf_hub_download
 
 class QalamNet:
     def __init__(self, repo_id="Ali0044/Qalam-Net"):
-        # 1. Download the portable model file from the Hub
-        print(f"Downloading model from {repo_id}...")
+        # 1. Download and Load Model
+        print(f"Loading Qalam-Net from {repo_id}...")
         model_path = hf_hub_download(repo_id=repo_id, filename="model.keras")
-        
-        # 2. Load the model (Self-contained v2 requires no custom_objects)
         self.model = keras.saving.load_model(model_path)
-        print("Model loaded successfully!")
         
-        # Standard Arabic Vocabulary
-        self.vocab = [' ', '!', '"', '#', '(', ')', '*', '+', ',', '-', '.', '/', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', ':', ';', '=', '?', '[', ']', 'ء', 'آ', 'أ', 'ؤ', 'إ', 'ئ', 'ا', 'ب', 'ة', 'ت', 'ث', 'ج', 'ح', 'خ', 'د', 'ذ', 'ر', 'ز', 'س', 'ش', 'ص', 'ض', 'ط', 'ظ', 'ع', 'غ', 'ـ', 'ف', 'ق', 'ك', 'ل', 'م', 'ن', 'ه', 'و', 'ى', 'ي', 'ً', 'ٌ', 'ٍ', 'َ', 'ُ', 'ِ', 'ّ', 'ْ', '٠', '١', '٢', '٣', '٤', '٥', '٦', '٧', '٨', '٩']
+        # 2. Define the exact 38-character Arabic Vocabulary
+        # [ALIF, BA, TA, THA, JEEM, HAA, KHAA, DAL, THAL, RA, ZAY, SEEN, SHEEN, SAD, DAD, TAA, ZAA, AIN, GHAIN, FA, QAF, KAF, LAM, MEEM, NOON, HA, WAW, YA, TEH_MARBUTA, ALEF_MAKSURA, ALEF_HAMZA_ABOVE, ALEF_HAMZA_BELOW, ALEF_MADDA, WAW_HAMZA, YEH_HAMZA, HAMZA, SPACE, TATWEEL]
+        self.vocab = ['ا', 'ب', 'ت', 'ث', 'ج', 'ح', 'خ', 'د', 'ذ', 'ر', 'ز', 'س', 'ش', 'ص', 'ض', 'ط', 'ظ', 'ع', 'غ', 'ف', 'ق', 'ك', 'ل', 'م', 'ن', 'ه', 'و', 'ي', 'ة', 'ى', 'أ', 'إ', 'آ', 'ؤ', 'ئ', 'ء', ' ', 'ـ']
         
     def preprocess(self, image_path):
         img = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
         img = cv2.resize(img, (128, 32)) / 255.0
-        img = img.T # (Height, Width) -> (Width, Height)
-        img = np.expand_dims(img, axis=(-1, 0)) # Add Channel and Batch
+        img = img.T # Transpose for CRNN architecture
+        img = np.expand_dims(img, axis=(-1, 0))
         return img.astype(np.float32)
 
     def predict(self, image_path):
         batch_img = self.preprocess(image_path)
-        predictions = self.model.predict(batch_img)
+        preds = self.model.predict(batch_img) # Output shape: (1, 32, 39)
+        
+        # 3. NumPy-based CTC Greedy Decoding (Cross-Backend)
+        argmax_preds = np.argmax(preds, axis=-1)[0]
+        
+        # Remove consecutive duplicates
+        unique_indices = [argmax_preds[i] for i in range(len(argmax_preds)) 
+                          if i == 0 or argmax_preds[i] != argmax_preds[i-1]]
         
-        # CTC Decode
-        input_len = np.ones(predictions.shape[0]) * predictions.shape[1]
-        results = keras.backend.ctc_decode(predictions, input_length=input_len, greedy=True)[0][0]
+        # Remove blank index (index 38)
+        blank_index = preds.shape[-1] - 1
+        final_indices = [idx for idx in unique_indices if idx != blank_index]
         
-        return "".join([self.vocab[int(res)] for res in results[0] if res != -1])
+        # Map to vocabulary
+        return "".join([self.vocab[idx] for idx in final_indices if idx < len(self.vocab)])
 
-# Run
+# Usage
 # ocr = QalamNet()
-# print(f"Output: {ocr.predict('test_sample.png')}")
+# print(f"Predicted Arabic Text: {ocr.predict('sample.png')}")
 ```
 
 ## 🧠 Model Architecture