market_w_tricks.yaml

index:
  feature_names: ['output']

  # dim_processors:
  #   names: ["L2Normalize"]

  dim_processors:
    names: ["L2Normalize", "PCA", "L2Normalize"]
    PCA:
      proj_dim: 560
      whiten: True
      train_fea_dir: "/home/data1/zgp/mmsport/2022-winners-player-reidentification-challenge-master/fea/test_vit_l_openai_colorjitetr_refea"
      l2: True


  feature_enhancer:
    name: "DBA"  # "Identity" / "DBA"
    DBA: 
      enhance_k: 5  # number of the nearest points to be calculated.

  metric:
    name: "KNN"
  
  re_ranker:
    name: "QE"  # "Identity"
    QE:
      qe_times: 1  # number of query expansion times.
      qe_k: 1 # number of the neighbors to be combined.

  # re_ranker:
  #   name: "KReciprocal"
  #   KReciprocal:
  #     k1: 20  # hyper-parameter for calculating jaccard distance.
  #     k2: 6  # hyper-parameter for calculating local query expansion.
  #     lambda_value: 0.3  # hyper-parameter for calculating the final distance.

  # re_ranker:
  #   name: "QEKR"
  #   QEKR:
  #     qe_times: 1  # number of query expansion times.
  #     qe_k: 1  # number of the neighbors to be combined.
  #     k1: 30  # hyper-parameter for calculating jaccard distance.    20->30->15
  #     k2: 3  # hyper-parameter for calculating local query expansion.   6->3->6
  #     lambda_value: 0.6  # hyper-parameter for calculating the final distance.