fix: handle ZeroDivisionError in embedding extraction

app.py

@@ -311,38 +311,82 @@ class FaceDetector:
 
     def extract_embeddings(self, face_img: np.ndarray) -> Tuple[Optional[np.ndarray], Optional[np.ndarray]]:
         try:
+            if face_img is None or face_img.size == 0:
+                logger.warning("Boş veya None face_img")
+                return None, None
+            
+            if face_img.shape[0] < 10 or face_img.shape[1] < 10:
+                logger.warning(f"Yüz çok küçük: {face_img.shape}")
+                return None, None
+            
             detector, recognizer = self.models
             bboxes, kpss = detector.autodetect(face_img, max_num=1)
             if len(bboxes) == 0:
+                logger.warning("Yüz tespit edilemedi")
                 return None, None
             kps = kpss[0]
+            if kps is None or len(kps) < 5:
+                logger.warning("Geçersiz keypoints")
+                return None, None
+            
             embedding = recognizer.get(face_img, kps)
+
+            if embedding is None:
+                logger.warning("Embedding None döndü")
+                return None, None
+            
+            if np.isnan(embedding).any() or np.isinf(embedding).any():
+                logger.warning("Embedding NaN veya Inf içeriyor")
+                return None, None            
+            
+            if np.allclose(embedding, 0):
+                logger.warning("Embedding sıfır vektör")
+                return None, None
+        
             return embedding, kps
         except Exception as e:
             logger.error(f"Embedding çıkarma hatası: {e}")
             return None, None
 
     def calculate_face_quality(self, face_img: np.ndarray, face_size: float, kps: np.ndarray) -> float:
-        quality_score = 0
-        size_score = min(face_size / 5000, 2.0)
-        quality_score += size_score
-        left_eye, right_eye = kps[0], kps[1]
-        eye_distance = np.linalg.norm(left_eye - right_eye)
-        face_width = np.sqrt(face_size)
-        eye_ratio = eye_distance / face_width
-        angle_score = min(eye_ratio * 3, 2.0)
-        quality_score += angle_score
-        gray = cv2.cvtColor(face_img, cv2.COLOR_BGR2GRAY)
-        blur_var = cv2.Laplacian(gray, cv2.CV_64F).var()
-        blur_score = min(blur_var / 500, 2.0)
-        quality_score += blur_score
-        left_mouth, right_mouth = kps[3], kps[4]
-        mouth_distance = np.linalg.norm(left_mouth - right_mouth)
-        mouth_ratio = mouth_distance / face_width
-        symmetry_score = min(mouth_ratio * 3, 2.0)
-        quality_score += symmetry_score
-        return quality_score
-
+        try:
+            quality_score = 0
+            if face_size <= 0:
+                logger.warning("Geçersiz face_size: <= 0")
+                return 0.0
+            
+            # Size score
+            size_score = min(face_size / 5000, 2.0)
+            quality_score += size_score
+            
+            # Keypoint kontrolü
+            if kps is None or len(kps) < 5:
+                logger.warning("Geçersiz keypoints")
+                return quality_score
+            
+            left_eye, right_eye = kps[0], kps[1]
+            eye_distance = np.linalg.norm(left_eye - right_eye)
+            face_width = np.sqrt(face_size)
+            if face_width <= 0:
+                logger.warning("face_width <= 0")
+                return quality_score
+            eye_ratio = eye_distance / face_width
+            angle_score = min(eye_ratio * 3, 2.0)
+            quality_score += angle_score
+            gray = cv2.cvtColor(face_img, cv2.COLOR_BGR2GRAY)
+            blur_var = cv2.Laplacian(gray, cv2.CV_64F).var()
+            blur_score = min(blur_var / 500, 2.0)
+            quality_score += blur_score
+            left_mouth, right_mouth = kps[3], kps[4]
+            mouth_distance = np.linalg.norm(left_mouth - right_mouth)
+            mouth_ratio = mouth_distance / face_width
+            symmetry_score = min(mouth_ratio * 3, 2.0)
+            quality_score += symmetry_score
+            return quality_score
+        except Exception as e:
+            logger.error(f"Quality hesaplama hatası: {e}")
+            return 1.0  
+        
     def process_frame(self, frame: np.ndarray) -> List[Dict]:
         frame = cv2.resize(frame, (0, 0), fx=self.config.resize_factor, fy=self.config.resize_factor)
         detector, _ = self.models

core/comparator.py

@@ -407,12 +407,10 @@ Her iki videoda da bulunan kişiler:
         
         report += f"""
 ### Video 1
-- **Yol**: {result.metadata['video1']['path']}
 - **Süre**: {result.metadata['video1']['duration']:.1f} saniye
 - **FPS**: {result.metadata['video1']['fps']}
 
 ### Video 2
-- **Yol**: {result.metadata['video2']['path']}
 - **Süre**: {result.metadata['video2']['duration']:.1f} saniye
 - **FPS**: {result.metadata['video2']['fps']}