inference_develop.yaml

# [Default Setting]
Default :
  RESULT_FOLDER: ./aicup_results
  FRAME_FOLDER: C:\AICUP\Tracking_dataset\valid\images
  # TIME_STAMP: 1016
  SELECT_DATE: 1016
  SELECT_TIME: 150000  # 150000 or 190000
  SELECT_CAMERA: all  # 0, 1, 2, 3, 4, 5, 6, 7, all

  # write mot (AICUP submit) txt file
  WRITE_MOT_TXT: true

  # write final inference video
  SAVE_OUT_VIDEO: false
  SAVE_OUT_VIDEO_FPS: 2


# [Detector]
Detector:
  ENSEMBLE: true  # if set true, fill the detector_weight_list and corresponding score
  ENSEMBLE_MODEL_LIST: [
        D:\Jonathan\AI_project\ObjectTracking\code\yolov9\runs\train\yolov9-c\weights\best.pt,
        D:\Jonathan\AI_project\ObjectTracking\code\yolov8_detector\weights\best.pt,
        D:\Jonathan\AI_project\ObjectTracking\code\yolov8_world\yolov8\yolov8x-worldv2.pt
    ]
  ENSEMBLE_WEIGHT_LIST: [0.7, 0.2, 0.1]

  DETECTOR_WEIGHT: D:\Jonathan\AI_project\ObjectTracking\code\yolov9\runs\train\yolov9-c\weights\best.pt
  DETECTOR_CONFIDENCE: 0.05


# [Extractor]
Extractor:
  EXTRACTOR_WEIGHT: D:\Jonathan\AI_project\ObjectTracking\code\yolov9\ReId\log\osnet_x1_0\model\model.pth.tar-50
  EXTRACTOR_TYPE: osnet_x1_0
  EXTRACTOR_THRESHOLD: 0.5


# [Tracker]
Tracker:
  TRACKER_MAX_UNMATCH_FRAME : 1