Update upload

app.py

@@ -82,8 +82,8 @@ def _load_model_background():
     _model_ready = True
     print("✅ Modèle YOLO chargé !")
 
-# ── CORRECTION 3 : utiliser yolov8n (plus stable sur HuggingFace Spaces) ─────
-def _get_model(key="yolov8n"):   # ← WAS "yolo11n" (moins fiable sur HF)
+# ── CORRECTION 3 : utiliser yolov11n (plus stable sur HuggingFace Spaces) ─────
+def _get_model(key="yolov11n"):   # ← WAS "yolo11n" (moins fiable sur HF)
     p = MODELS_DIR / f"{key}.pt"
     if not p.exists():
         m = YOLO(f"{key}.pt")
@@ -248,6 +248,7 @@ def _worker(jid):
     print(f"[PROCESS] Starting frame processing...\n")
     
     frame_count = 0
+    frame_detections = []  # Store all detections for CSV export
     
     while True:
         ret, frame = cap.read()
@@ -255,6 +256,7 @@ def _worker(jid):
             break
         
         frame_count += 1
+        timestamp_sec = (frame_count - 1) / fps if fps > 0 else 0
         
         # Run inference
         results = _model(frame, conf=CONF, iou=IOU, imgsz=INFER_SZ, verbose=False)
@@ -270,8 +272,35 @@ def _worker(jid):
                 if class_name:
                     job["detections"][class_name] = job["detections"].get(class_name, 0) + 1
                     
-                    # Draw bounding box
+                    # Get bounding box coordinates
                     x1, y1, x2, y2 = map(int, box.xyxy[0])
+                    cx = (x1 + x2) // 2
+                    cy = (y1 + y2) // 2
+                    
+                    # Store detection data for CSV
+                    frame_detections.append({
+                        "frame": frame_count,
+                        "timestamp_sec": timestamp_sec,
+                        "scene_name": jid,
+                        "group_id": jid,
+                        "video_name": job["name"],
+                        "track_id": "",
+                        "class_name": class_name,
+                        "confidence": conf,
+                        "bbox_x1": x1,
+                        "bbox_y1": y1,
+                        "bbox_x2": x2,
+                        "bbox_y2": y2,
+                        "cx": cx,
+                        "cy": cy,
+                        "frame_width": width,
+                        "frame_height": height,
+                        "crossed_line": "false",
+                        "direction": "",
+                        "speed_px_s": 0.0
+                    })
+                    
+                    # Draw bounding box
                     color = CLASS_COLORS.get(class_name, (255, 255, 255))
                     label = f"{class_name} {conf:.2f}"
                     
@@ -288,6 +317,9 @@ def _worker(jid):
             detections = sum(job['detections'].values())
             print(f"[PROCESS] Frame {frame_count}/{total_frames} ({progress:.1f}%) - Detections: {detections}")
     
+    # Save frame detections for later CSV export
+    job["frame_detections"] = frame_detections
+    
     # Release resources
     cap.release()
     out.release()
@@ -359,11 +391,40 @@ def api_logs_csv(scene_id):
 
     output = io.StringIO()
     writer = csv.writer(output)
-    writer.writerow(["scene_id", "video_name", "class_name", "count",
-                     "generated_at", "total_frames"])
-    for class_name, count in job.get("detections", {}).items():
-        writer.writerow([scene_id, job["name"], class_name, count,
-                         datetime.now().isoformat(), job.get("frames", 0)])
+    
+    # Write CSV header - SCHEMA_EXAMPLE.csv format
+    writer.writerow([
+        "frame", "timestamp_sec", "scene_name", "group_id", 
+        "video_name", "track_id", "class_name", "confidence",
+        "bbox_x1", "bbox_y1", "bbox_x2", "bbox_y2", 
+        "cx", "cy", "frame_width", "frame_height",
+        "crossed_line", "direction", "speed_px_s"
+    ])
+    
+    # Write detections for each frame
+    frame_detections = job.get("frame_detections", [])
+    for det in frame_detections:
+        writer.writerow([
+            det["frame"],
+            f"{det['timestamp_sec']:.3f}",
+            det["scene_name"],
+            det["group_id"],
+            det["video_name"],
+            det["track_id"],
+            det["class_name"],
+            f"{det['confidence']:.3f}",
+            det["bbox_x1"],
+            det["bbox_y1"],
+            det["bbox_x2"],
+            det["bbox_y2"],
+            det["cx"],
+            det["cy"],
+            det["frame_width"],
+            det["frame_height"],
+            det["crossed_line"],
+            det["direction"],
+            f"{det['speed_px_s']:.1f}"
+        ])
 
     return Response(
         output.getvalue(),

test_csv_format.py

@@ -0,0 +1,140 @@
+#!/usr/bin/env python3
+"""
+Test script to verify CSV format matches SCHEMA_EXAMPLE.csv
+"""
+
+import csv
+from io import StringIO
+
+# Exemple de détections comme generées par le backend
+frame_detections = [
+    {
+        "frame": 1,
+        "timestamp_sec": 0.033,
+        "scene_name": "abc123",
+        "group_id": "abc123",
+        "video_name": "traffic_clip1.mp4",
+        "track_id": "",
+        "class_name": "Vehicle",
+        "confidence": 0.912,
+        "bbox_x1": 340,
+        "bbox_y1": 210,
+        "bbox_x2": 520,
+        "bbox_y2": 310,
+        "cx": 430,
+        "cy": 260,
+        "frame_width": 1920,
+        "frame_height": 1080,
+        "crossed_line": "false",
+        "direction": "",
+        "speed_px_s": 0.0
+    },
+    {
+        "frame": 2,
+        "timestamp_sec": 0.067,
+        "scene_name": "abc123",
+        "group_id": "abc123",
+        "video_name": "traffic_clip1.mp4",
+        "track_id": "",
+        "class_name": "Vehicle",
+        "confidence": 0.908,
+        "bbox_x1": 345,
+        "bbox_y1": 213,
+        "bbox_x2": 525,
+        "bbox_y2": 313,
+        "cx": 435,
+        "cy": 263,
+        "frame_width": 1920,
+        "frame_height": 1080,
+        "crossed_line": "false",
+        "direction": "",
+        "speed_px_s": 14.2
+    },
+    {
+        "frame": 48,
+        "timestamp_sec": 1.600,
+        "scene_name": "abc123",
+        "group_id": "abc123",
+        "video_name": "traffic_clip1.mp4",
+        "track_id": "",
+        "class_name": "Vehicle",
+        "confidence": 0.887,
+        "bbox_x1": 560,
+        "bbox_y1": 290,
+        "bbox_x2": 740,
+        "bbox_y2": 390,
+        "cx": 650,
+        "cy": 340,
+        "frame_width": 1920,
+        "frame_height": 1080,
+        "crossed_line": "true",
+        "direction": "down",
+        "speed_px_s": 15.1
+    },
+]
+
+# Generate CSV
+output = StringIO()
+writer = csv.writer(output)
+
+# Header
+writer.writerow([
+    "frame", "timestamp_sec", "scene_name", "group_id",
+    "video_name", "track_id", "class_name", "confidence",
+    "bbox_x1", "bbox_y1", "bbox_x2", "bbox_y2",
+    "cx", "cy", "frame_width", "frame_height",
+    "crossed_line", "direction", "speed_px_s"
+])
+
+# Data rows
+for det in frame_detections:
+    writer.writerow([
+        det["frame"],
+        f"{det['timestamp_sec']:.3f}",
+        det["scene_name"],
+        det["group_id"],
+        det["video_name"],
+        det["track_id"],
+        det["class_name"],
+        f"{det['confidence']:.3f}",
+        det["bbox_x1"],
+        det["bbox_y1"],
+        det["bbox_x2"],
+        det["bbox_y2"],
+        det["cx"],
+        det["cy"],
+        det["frame_width"],
+        det["frame_height"],
+        det["crossed_line"],
+        det["direction"],
+        f"{det['speed_px_s']:.1f}"
+    ])
+
+# Print result
+csv_output = output.getvalue()
+print("Generated CSV:")
+print("=" * 120)
+print(csv_output)
+print("=" * 120)
+
+# Compare with expected format
+expected_header = "frame,timestamp_sec,scene_name,group_id,video_name,track_id,class_name,confidence,bbox_x1,bbox_y1,bbox_x2,bbox_y2,cx,cy,frame_width,frame_height,crossed_line,direction,speed_px_s"
+actual_header = csv_output.split('\n')[0]
+
+print("\nExpected header:")
+print(expected_header)
+print("\nActual header:")
+print(actual_header)
+print("\nHeaders match:", expected_header == actual_header)
+
+# Verify number of columns
+lines = csv_output.strip().split('\n')
+for i, line in enumerate(lines, 1):
+    cols = line.split(',')
+    expected_cols = 19
+    if len(cols) != expected_cols:
+        print(f"Row {i}: WARNING - Expected {expected_cols} columns, got {len(cols)}")
+    else:
+        print(f"Row {i}: OK ({len(cols)} columns)")
+
+print("\n✅ CSV Format validation complete!")