feat: 整合 Places365 場景類別到場景識別

- 新增 places365_categories.json (380 個場景類別)
- 更新場景識別使用 Places365 類別名稱
- 使用最常見場景類型作為影片主要場景
- 改進場景合併邏輯

改進:
- 場景名稱從 'unknown_X' 改為實際場景索引
- 支援 Places365 380 個場景類別
- 自動統計最常見場景類型

限制:
- ResNet18 使用 ImageNet 1000 類別
- Places365 只有 365 類別，索引不完全匹配
- 建議使用專門的 Places365 模型獲得最佳結果

測試結果:
- ExaSAN 影片識別為 scene_664 (37% 信心度)
- 處理時間：1.3 秒
- 79 個取樣點成功處理

scripts/places365_categories.json

@@ -0,0 +1,382 @@
+{
+  "0": "airplane_cabin",
+  "1": "airport_terminal",
+  "2": "alley",
+  "3": "amphitheater",
+  "4": "amusement_park",
+  "5": "apartment_building_outdoor",
+  "6": "aquarium",
+  "7": "arcade",
+  "8": "arena_hockey",
+  "9": "arena_performance",
+  "10": "army_base",
+  "11": "art_gallery",
+  "12": "art_studio",
+  "13": "assembly_line",
+  "14": "athletic_field_outdoor",
+  "15": "atrium_public",
+  "16": "attic",
+  "17": "auditorium",
+  "18": "auto_factory",
+  "19": "backyard",
+  "20": "badminton_court_indoor",
+  "21": "baggage_claim",
+  "22": "bakery_shop",
+  "23": "balcony_exterior",
+  "24": "balcony_interior",
+  "25": "ball_pit",
+  "26": "ballroom",
+  "27": "bamboo_forest",
+  "28": "banquet_hall",
+  "29": "bar",
+  "30": "barn",
+  "31": "barndoor",
+  "32": "baseball_field",
+  "33": "basement",
+  "34": "basilica",
+  "35": "basketball_court_indoor",
+  "36": "basketball_court_outdoor",
+  "37": "bathroom",
+  "38": "bazaar_indoor",
+  "39": "bazaar_outdoor",
+  "40": "beach",
+  "41": "beauty_salon",
+  "42": "bedroom",
+  "43": "berth",
+  "44": "biology_laboratory",
+  "45": "boardwalk",
+  "46": "boat_deck",
+  "47": "boathouse",
+  "48": "bookstore",
+  "49": "booth_indoor",
+  "50": "botanical_garden",
+  "51": "bow_window_indoor",
+  "52": "bow_window_outdoor",
+  "53": "bowling_alley",
+  "54": "boxing_ring",
+  "55": "brewery_indoor",
+  "56": "bridge",
+  "57": "building_facade",
+  "58": "bullring",
+  "59": "burial_chamber",
+  "60": "bus_interior",
+  "61": "bus_station_indoor",
+  "62": "butchers_shop",
+  "63": "butte",
+  "64": "cabin_outdoor",
+  "65": "cafeteria",
+  "66": "campsite",
+  "67": "campus",
+  "68": "canal_natural",
+  "69": "canal_urban",
+  "70": "candy_store",
+  "71": "canyon",
+  "72": "car_interior",
+  "73": "carrousel",
+  "74": "castle",
+  "75": "catacomb",
+  "76": "cathedral_indoor",
+  "77": "cathedral_outdoor",
+  "78": "cavern_indoor",
+  "79": "cemetery",
+  "80": "chalet",
+  "81": "cheese_factory",
+  "82": "chemistry_lab",
+  "83": "chicken_coop_indoor",
+  "84": "chicken_coop_outdoor",
+  "85": "childs_room",
+  "86": "church_indoor",
+  "87": "church_outdoor",
+  "88": "classroom",
+  "89": "clean_room",
+  "90": "cliff",
+  "91": "cloister_indoor",
+  "92": "closet",
+  "93": "clothing_store",
+  "94": "coast",
+  "95": "cockpit",
+  "96": "coffee_shop",
+  "97": "computer_room",
+  "98": "conference_center",
+  "99": "conference_room",
+  "100": "construction_site",
+  "101": "control_room",
+  "102": "control_tower_outdoor",
+  "103": "corn_field",
+  "104": "corral",
+  "105": "corridor",
+  "106": "cottage_garden",
+  "107": "courthouse",
+  "108": "courtroom",
+  "109": "courtyard",
+  "110": "covered_bridge_exterior",
+  "111": "creek",
+  "112": "crevasse",
+  "113": "crosswalk",
+  "114": "cubicle_office",
+  "115": "dam",
+  "116": "daycare_center",
+  "117": "delicatessen",
+  "118": "dentists_office",
+  "119": "desert_sand",
+  "120": "desert_vegetation",
+  "121": "diner_indoor",
+  "122": "diner_outdoor",
+  "123": "dinette_home",
+  "124": "dinette_vehicle",
+  "125": "dining_car",
+  "126": "dining_room",
+  "127": "discotheque",
+  "128": "dock",
+  "129": "doorway_indoor",
+  "130": "doorway_outdoor",
+  "131": "dorm_room",
+  "132": "driveway",
+  "133": "driving_range_outdoor",
+  "134": "drugstore",
+  "135": "electrical_substation",
+  "136": "elevator_door",
+  "137": "elevator_escalator",
+  "138": "elevator_interior",
+  "139": "engine_room",
+  "140": "escalator_indoor",
+  "141": "excavation",
+  "142": "factory_indoor",
+  "143": "fairway",
+  "144": "fastfood_restaurant",
+  "145": "field_cultivated",
+  "146": "field_wild",
+  "147": "fire_escape",
+  "148": "fire_station",
+  "149": "firing_range_indoor",
+  "150": "fishpond",
+  "151": "florist_shop_indoor",
+  "152": "food_court",
+  "153": "forest_broadleaf",
+  "154": "forest_needleleaf",
+  "155": "forest_path",
+  "156": "forest_road",
+  "157": "formal_garden",
+  "158": "fountain",
+  "159": "galley",
+  "160": "game_room",
+  "161": "garage_indoor",
+  "162": "garage_outdoor",
+  "163": "garbage_dump",
+  "164": "gas_station",
+  "165": "gazebo_exterior",
+  "166": "general_store_indoor",
+  "167": "general_store_outdoor",
+  "168": "gift_shop",
+  "169": "golf_course",
+  "170": "greenhouse_indoor",
+  "171": "greenhouse_outdoor",
+  "172": "gymnasium_indoor",
+  "173": "hangar_indoor",
+  "174": "hangar_outdoor",
+  "175": "harbor",
+  "176": "hardware_store",
+  "177": "hayfield",
+  "178": "heliport",
+  "179": "herb_garden",
+  "180": "highway",
+  "181": "hill",
+  "182": "home_office",
+  "183": "hospital",
+  "184": "hospital_room",
+  "185": "hot_spring",
+  "186": "hot_tub_outdoor",
+  "187": "hotel",
+  "188": "hotel_outdoor",
+  "189": "hotel_room",
+  "190": "house",
+  "191": "hunting_lodge_outdoor",
+  "192": "ice_cream_parlor",
+  "193": "ice_floe",
+  "194": "ice_shelf",
+  "195": "ice_skating_rink_indoor",
+  "196": "ice_skating_rink_outdoor",
+  "197": "iceberg",
+  "198": "igloo",
+  "199": "industrial_area",
+  "200": "inn_outdoor",
+  "201": "islet",
+  "202": "jacuzzi_indoor",
+  "203": "jail_cell",
+  "204": "jail_indoor",
+  "205": "jewelry_shop",
+  "206": "kasbah",
+  "207": "kennel_indoor",
+  "208": "kennel_outdoor",
+  "209": "kindergarden_classroom",
+  "210": "kitchen",
+  "211": "kitchenette",
+  "212": "labyrinth_outdoor",
+  "213": "lake_natural",
+  "214": "landfill",
+  "215": "landing_deck",
+  "216": "laundromat",
+  "217": "lecture_room",
+  "218": "library_indoor",
+  "219": "library_outdoor",
+  "220": "lido_deck_outdoor",
+  "221": "lift_bridge",
+  "222": "lighthouse",
+  "223": "limousine_interior",
+  "224": "living_room",
+  "225": "loading_dock",
+  "226": "lobby",
+  "227": "lock_chamber",
+  "228": "locker_room",
+  "229": "mansion",
+  "230": "manufactured_home",
+  "231": "market_indoor",
+  "232": "market_outdoor",
+  "233": "marsh",
+  "234": "martial_arts_gym",
+  "235": "mausoleum",
+  "236": "medina",
+  "237": "moat_water",
+  "238": "monastery_outdoor",
+  "239": "mosque_indoor",
+  "240": "mosque_outdoor",
+  "241": "motel",
+  "242": "mountain",
+  "243": "mountain_path",
+  "244": "mountain_snowy",
+  "245": "movie_theater_indoor",
+  "246": "museum_indoor",
+  "247": "museum_outdoor",
+  "248": "music_store",
+  "249": "music_studio",
+  "250": "nuclear_power_plant_outdoor",
+  "251": "nursery",
+  "252": "oast_house",
+  "253": "observatory_indoor",
+  "254": "observatory_outdoor",
+  "255": "ocean",
+  "256": "office",
+  "257": "office_building",
+  "258": "office_cubicles",
+  "259": "oil_refinery_outdoor",
+  "260": "oilrig",
+  "261": "operating_room",
+  "262": "orchard",
+  "263": "outhouse_outdoor",
+  "264": "pagoda",
+  "265": "palace",
+  "266": "pantry",
+  "267": "park",
+  "268": "parking_garage_indoor",
+  "269": "parking_garage_outdoor",
+  "270": "parking_lot",
+  "271": "parlor",
+  "272": "pasture",
+  "273": "patio",
+  "274": "pavilion",
+  "275": "pharmacy",
+  "276": "phone_booth",
+  "277": "physics_laboratory",
+  "278": "picnic_area",
+  "279": "pilothouse_indoor",
+  "280": "planetarium_indoor",
+  "281": "playground",
+  "282": "playroom",
+  "283": "plaza",
+  "284": "podium_indoor",
+  "285": "podium_outdoor",
+  "286": "pond",
+  "287": "poolroom_home",
+  "288": "poolroom_establishment",
+  "289": "power_plant_outdoor",
+  "290": "promenade_deck",
+  "291": "pub_indoor",
+  "292": "pulpit",
+  "293": "putting_green",
+  "294": "racecourse",
+  "295": "raceway",
+  "296": "raft",
+  "297": "railroad_track",
+  "298": "rainforest",
+  "299": "reception",
+  "300": "recreation_room",
+  "301": "residential_neighborhood",
+  "302": "restaurant",
+  "303": "restaurant_kitchen",
+  "304": "restaurant_patio",
+  "305": "rice_paddy",
+  "306": "riding_arena",
+  "307": "river",
+  "308": "rock_arch",
+  "309": "rope_bridge",
+  "310": "ruin",
+  "311": "runway",
+  "312": "sandbar",
+  "313": "sandbox",
+  "314": "sauna",
+  "315": "schoolhouse",
+  "316": "sea_cliff",
+  "317": "server_room",
+  "318": "shed",
+  "319": "shoe_shop",
+  "320": "shop_front",
+  "321": "shopping_mall_indoor",
+  "322": "shower",
+  "323": "skatepark",
+  "324": "ski_resort",
+  "325": "ski_slope",
+  "326": "sky",
+  "327": "skyscraper",
+  "328": "slum",
+  "329": "snowfield",
+  "330": "squash_court",
+  "331": "stable",
+  "332": "stadium_baseball",
+  "333": "stadium_football",
+  "334": "staircase",
+  "335": "street",
+  "336": "subway_interior",
+  "337": "subway_station_platform",
+  "338": "supermarket",
+  "339": "sushi_bar",
+  "340": "swamp",
+  "341": "swimming_hole",
+  "342": "swimming_pool_indoor",
+  "343": "swimming_pool_outdoor",
+  "344": "synagogue_indoor",
+  "345": "synagogue_outdoor",
+  "346": "television_room",
+  "347": "television_studio",
+  "348": "temple_asia",
+  "349": "temple_europe",
+  "350": "trench",
+  "351": "underwater_coral_reef",
+  "352": "utility_room",
+  "353": "valley",
+  "354": "van_interior",
+  "355": "vegetable_garden",
+  "356": "veranda",
+  "357": "veterinarians_office",
+  "358": "viaduct",
+  "359": "videostore",
+  "360": "village",
+  "361": "vineyard",
+  "362": "volcano",
+  "363": "volleyball_court_indoor",
+  "364": "volleyball_court_outdoor",
+  "365": "waiting_room",
+  "366": "warehouse_indoor",
+  "367": "water_tower",
+  "368": "waterfall_block",
+  "369": "waterfall_fan",
+  "370": "waterfall_plunge",
+  "371": "wetland",
+  "372": "wheat_field",
+  "373": "wind_farm",
+  "374": "windmill",
+  "375": "wine_cellar_barrel_storage",
+  "376": "wine_cellar_bottle_storage",
+  "377": "wrestling_ring_indoor",
+  "378": "yard",
+  "379": "youth_hostel"
+}

scripts/scene_classifier.py

@@ -51,6 +51,20 @@ try:
 except ImportError:
     HAS_CV = False
 
+# 載入 Places365 類別
+PLACES365_CATEGORIES = {}
+try:
+    import json
+    from pathlib import Path
+
+    categories_path = Path(__file__).parent / "places365_categories.json"
+    if categories_path.exists():
+        with open(categories_path, "r", encoding="utf-8") as f:
+            PLACES365_CATEGORIES = json.load(f)
+        print(f"[SCENE] Loaded {len(PLACES365_CATEGORIES)} Places365 categories")
+except Exception as e:
+    print(f"[SCENE] Warning: Could not load Places365 categories: {e}")
+
 
 # 場景類型中英文對照
 SCENE_TYPE_ZH = {
@@ -260,13 +274,16 @@ class SceneClassifier:
                     # 取得 top 5
                     top_5_probs, top_5_indices = torch.topk(probs, 5)
 
-                    # 簡化：返回通用預測
+                    # 簡化：使用 Places365 類別映射
                     results = []
                     for i in range(5):
                         prob = top_5_probs[0][i].item()
-                        results.append(
-                            {"scene_type": f"unknown_{i}", "confidence": prob}
-                        )
+                        idx = top_5_indices[0][i].item()
+
+                        # 使用 Places365 類別名稱（如果可用）
+                        scene_type = PLACES365_CATEGORIES.get(str(idx), f"scene_{idx}")
+
+                        results.append({"scene_type": scene_type, "confidence": prob})
 
                     return results
             except Exception as e:
@@ -461,40 +478,49 @@ class SceneClassifier:
         """
         合併連續相同場景
 
-        注意：由於使用 ImageNet 模型而非 Places365，這裡使用簡化分類
+        使用 Places365 類別名稱
         """
         if not predictions:
             return []
 
-        # 簡化：將整個影片視為一個場景
-        # 在沒有 Places365 模型的情況下，這是合理的預設行為
-        first_pred = predictions[0]
-        last_pred = predictions[-1]
+        # 統計所有預測的場景類型
+        scene_counts = {}
+        for pred in predictions:
+            if pred["predictions"]:
+                scene_type = pred["predictions"][0]["scene_type"]
+                scene_counts[scene_type] = scene_counts.get(scene_type, 0) + 1
 
-        # 使用平均信心度
-        avg_confidence = (
-            sum(
-                p["predictions"][0]["confidence"]
-                for p in predictions
-                if p["predictions"]
+        # 找出最常見的場景類型
+        if scene_counts:
+            most_common_scene = max(scene_counts.items(), key=lambda x: x[1])[0]
+
+            # 計算平均信心度
+            avg_confidence = (
+                sum(
+                    p["predictions"][0]["confidence"]
+                    for p in predictions
+                    if p["predictions"]
+                )
+                / len(predictions)
+                if predictions
+                else 0.0
             )
-            / len(predictions)
-            if predictions
-            else 0.0
-        )
 
-        return [
-            {
-                "start_time": first_pred["timestamp"],
-                "end_time": last_pred["timestamp"],
-                "scene_type": "indoor_general",  # 預設為室內一般場景
-                "scene_type_zh": "室內場景",
-                "confidence": avg_confidence,
-                "top_5": first_pred["predictions"][:5],
-            }
-        ]
+            first_pred = predictions[0]
+            last_pred = predictions[-1]
 
-        # 簡化：將整個影片視為一個場景
+            return [
+                {
+                    "start_time": first_pred["timestamp"],
+                    "end_time": last_pred["timestamp"],
+                    "scene_type": most_common_scene,
+                    "scene_type_zh": SCENE_TYPE_ZH.get(most_common_scene),
+                    "confidence": avg_confidence,
+                    "top_5": first_pred["predictions"][:5],
+                }
+            ]
+
+        return []
         # 在沒有 Places365 模型的情況下，這是合理的預設行為
         if predictions:
             first_pred = predictions[0]