Experiment Science and Technology, Volume. 22, Issue 1, 22(2024)

Fusion Experimental Method of Soft Biometrics Based on Personalized Least Squares

Qing ZHANG and Yangyang ZHANG*
Author Affiliations
  • Fine Arts School of Shandong University, Jinan 250100, China
  • show less
    Figures & Tables(15)
    • Table 1. [in Chinese]

      View table
      View in Article

      Table 1. [in Chinese]

      用户软特征PCALDA
      δ 集成主特征的识别性能 EER/% δ 集成主特征的识别性能 EER/%
      用户11号0.780.1140.750.078
      2号0.850.1050.810.072
      3号0.670.1350.640.086
      4号0.860.1070.900.064
      5号0.730.1250.800.074
      用户21号0.660.1350.710.083
      2号0.750.1170.810.075
      3号0.780.1120.850.072
      4号0.860.1010.900.066
      5号0.690.1280.700.083
    • Table 2. [in Chinese]

      View table
      View in Article

      Table 2. [in Chinese]

      用户软特征PCALDA
      δ 集成主特征的识别性能 EER/% δ 集成主特征的识别性能 EER/%
      用户11号0.710.0280.680.041
      2号0.640.0360.650.041
      3号0.850.0220.880.033
      4号0.690.0310.870.035
      5号0.750.0250.700.038
      用户21号0.750.0230.770.034
      2号0.850.0190.860.035
      3号0.640.0380.680.042
      4号0.910.0160.850.035
      5号0.790.0220.700.037
    • Table 3. [in Chinese]

      View table
      View in Article

      Table 3. [in Chinese]

      数据库基方法EER/%
      本文集成 方法文献[9]集成方法不集成软特征的方法只使用软特征的方法
      FacePixPCA0.080.100.1411.65
      LDA0.040.050.0913.33
      ICA0.090.120.1714.58
      LLE0.080.110.1914.32
      Face-MLAPCA0.010.010.047.86
      LDA0.020.040.057.43
      ICA0.020.030.048.72
      LLE0.010.030.089.55
    • Table 4. [in Chinese]

      View table
      View in Article

      Table 4. [in Chinese]

      数据库基方法FAR/%
      本文集成 方法文献[9]集成方法不集成软特征的方法只使用软特征的方法
      FacePixPCA0.310.370.3983.45
      LDA0.680.830.9879.92
      ICA0.400.490.5384.95
      LLE0.250.350.4788.96
      Face- MLAPCA0.020.020.0571.75
      LDA0.050.080.0967.83
      ICA0.040.050.0873.81
      LLE0.050.090.1678.85
    • Table 5. [in Chinese]

      View table
      View in Article

      Table 5. [in Chinese]

      数据库基方法FRR/%
      本文集成方法文献[9]集成方法不集成软特征的方法只使用软特征的方法
      FacePixPCA0.180.250.4359.50
      LDA0.240.350.6565.97
      ICA0.210.290.4268.89
      LLE0.310.390.4666.92
      Face- MLAPCA0.020.030.1157.85
      LDA0.050.090.2960.89
      ICA0.040.050.1766.91
      LLE0.070.100.2151.88
    • Table 6. [in Chinese]

      View table
      View in Article

      Table 6. [in Chinese]

      用户软特征DB1DB2
      δ 集成主特征的识别性能 EER/% δ 集成主特征的识别性能 EER/%
      用户1软特征I0.742.680.710.75
      软特征II0.821.830.680.79
      软特征III0.792.570.830.68
      用户2软特征I0.681.890.720.77
      软特征II0.831.770.670.83
      软特征III0.652.340.800.66
    • Table 7. [in Chinese]

      View table
      View in Article

      Table 7. [in Chinese]

      用户软特征DB1DB2
      δ 集成主特征的识别性能 EER/% δ 集成主特征的识别性能 EER/%
      用户1软特征I0.671.210.731.21
      软特征II0.810.980.841.03
      软特征III0.791.190.761.16
      用户2软特征I0.641.270.721.07
      软特征II0.830.950.870.95
      软特征III0.811.020.681.28
    • Table 8. [in Chinese]

      View table
      View in Article

      Table 8. [in Chinese]

      用户软特征DB1DB2
      δ 集成主特征的识别性能 EER/% δ 集成主特征的识别性能 EER/%
      用户1软特征I0.682.870.783.25
      软特征II0.891.940.852.88
      软特征III0.752.640.713.40
      用户2软特征I0.682.580.743.15
      软特征II0.842.210.803.05
      软特征III0.762.460.693.65
    • Table 9. [in Chinese]

      View table
      View in Article

      Table 9. [in Chinese]

      方法评价指标子库
      DB1DB2DB3DB4
      本文集成方法EER1.330.522.010.41
      FAR(FRR=0%)16.1426.8725.140.73
      FRR(FAR=0%)3.940.407.051.97
      文献[16]集成 方法EER1.500.672.620.58
      FAR(FRR=0%)19.0730.1827.840.83
      FRR(FAR=0%)6.001.0010.252.24
      不集成软特征的方法EER3.381.016.511.32
      FAR(FRR=0%)64.1066.0550.9811.90
      FRR(FAR=0%)18.005.2521.7514.42
      软特征IEER6.283.048.814.12
      FAR(FRR=0%)78.6885.4474.3566.36
      FRR(FAR=0%)43.3534.5645.1736.87
      软特征IIEER5.924.4510.937.01
      FAR(FRR=0%)73.3879.9683.9580.04
      FRR(FAR=0%)44.6844.6556.7073.44
      软特征IIIEER5.097.0716.447.00
      FAR(FRR=0%)70.2680.7887.6582.36
      FRR(FAR=0%)52.7967.7357.9479.60
    • Table 10. [in Chinese]

      View table
      View in Article

      Table 10. [in Chinese]

      方法评价指标子库
      DB1DB2DB3DB4
      本文方法EER0.380.540.560.30
      FAR(FRR=0%)10.9537.811.020.31
      FRR(FAR=0%)1.261.302.200.84
      文献[16] 方法EER0.500.730.760.33
      FAR(FRR=0%)14.2143.6714.110.57
      FRR(FAR=0%)1.501.502.751.25
      不集成软特征的方法EER1.471.383.311.37
      FAR(FRR=0%)41.7774.6279.649.33
      FRR(FAR=0%)4.254.0018.006.00
      软特征IEER4.492.464.534.12
      FAR(FRR=0%)89.5775.6584.3856.17
      FRR(FAR=0%)44.7952.6768.5469.31
      软特征IIEER5.924.4510.937.01
      FAR(FRR=0%)84.9083.3382.6771.79
      FRR(FAR=0%)57.8868.4788.9874.14
      软特征IIIEER4.572.2516.057.00
      FAR(FRR=0%)86.5971.3388.9081.27
      FRR(FAR=0%)53.1356.3567.8563.54
    • Table 11. [in Chinese]

      View table
      View in Article

      Table 11. [in Chinese]

      方法评价指标子库
      DB1DB2DB3DB4
      本文方法EER0.981.853.050.42
      FAR(FRR=0%)14.9534.3138.701.15
      FRR(FAR=0%)3.974.2614.301.95
      文献[16] 方法EER1.322.203.640.62
      FAR(FRR=0%)19.0745.4250.701.92
      FRR(FAR=0%)6.007.0016.752.25
      不集成软特征的方法EER3.394.107.941.61
      FAR(FRR=0%)64.1068.9968.4823.37
      FRR(FAR=0%)18.0012.7530.5011.25
      软特征IEER7.236.519.749.50
      FAR(FRR=0%)84.3389.3584.3576.33
      FRR(FAR=0%)49.3444.5359.1867.87
      软特征IIEER11.698.349.6110.71
      FAR(FRR=0%)89.9683.2588.9090.33
      FRR(FAR=0%)76.5467.5371.3369.68
      软特征IIIEER7.6315.9910.4414.68
      FAR(FRR=0%)83.2592.3387.7791.35
      FRR(FAR=0%)52.3371.0770.8578.67
    Tools

    Get Citation

    Copy Citation Text

    Qing ZHANG, Yangyang ZHANG. Fusion Experimental Method of Soft Biometrics Based on Personalized Least Squares[J]. Experiment Science and Technology, 2024, 22(1): 22

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category:

    Received: Nov. 3, 2022

    Accepted: --

    Published Online: Mar. 27, 2024

    The Author Email: ZHANG Yangyang (张洋洋)

    DOI:10.12179/1672-4550.20220630

    Topics