• Laser & Optoelectronics Progress
  • Vol. 57, Issue 22, 221012 (2020)
Wei Lü1, Desheng Li1、*, Lang Tan2, Peiguang Jing1, and Yuting Su1
Author Affiliations
  • 1School of Electronics and Information Engineering, Tianjin University, Tianjin 300072, China
  • 2Beijing Smartchip Microelectronics Technology Co., Ltd., Beijing 102200, China
  • show less
    DOI: 10.3788/LOP57.221012 Cite this Article Set citation alerts
    Wei Lü, Desheng Li, Lang Tan, Peiguang Jing, Yuting Su. Microvideo Multilabel Learning Model Based on Multiview Low-Rank Representation[J]. Laser & Optoelectronics Progress, 2020, 57(22): 221012 Copy Citation Text show less
    Illustration of proposed model
    Fig. 1. Illustration of proposed model
    Sample video with different labels selected from dataset
    Fig. 2. Sample video with different labels selected from dataset
    Convergence verification graphs. (a) Variation of Zdiff with model iteration; (b) variation of average precision with model iteration
    Fig. 3. Convergence verification graphs. (a) Variation of Zdiff with model iteration; (b) variation of average precision with model iteration
    Effect of different parameters on average precision. (a) Effect of λ2 on average precision; (b) effect of λ4 on average precision
    Fig. 4. Effect of different parameters on average precision. (a) Effect of λ2 on average precision; (b) effect of λ4 on average precision
    Label correlation matrix comparison. (a) Normalized correlation matrix for true label; (b) normalized correlation matrix S˙after the iteration
    Fig. 5. Label correlation matrix comparison. (a) Normalized correlation matrix for true label; (b) normalized correlation matrix S˙after the iteration
    Evaluation metricsNo LRNo LFNo LC
    Average precision difference-0.0508-0.2423-0.0172
    Hamming loss difference0.00130.00290.0005
    Ranking loss difference0.00790.05210.0049
    Coverage difference0.34924.69140.2851
    One-error difference0.01750.14820.0230
    Table 1. Ablation experiment results
    MethodAverage precisionHamming lossRanking lossCoverageOne-error
    DNMF0.4673±0.00630.0154±0.00010.1077±0.00828.3853±0.16210.6487±0.0082
    LRR0.5489±0.00570.0154±0.00010.0991±0.00518.4056±0.18030.3039±0.0057
    GLOCAL0.7527±0.00640.0133±0.00200.0515±0.00153.9943±0.10560.2457±0.0032
    MLKNN0.7843±0.00530.0134±0.00010.0476±0.00584.0204±0.18740.3087±0.0058
    Googlenet0.6676±0.00440.0176±0.00020.4349±0.00664.5680±0.06000.4349±0.0066
    C3D0.7149±0.00890.0146±0.00030.3694±0.00283.9041±0.20330.3694±0.0088
    C2AE0.8013±0.00220.0128±0.00010.0481±0.00413.6942±0.14710.2381±0.0026
    Proposed0.8055±0.00280.0128±0.00010.0432±0.00233.6732±0.12740.2561±0.0065
    Table 2. Performance comparison of different algorithms
    Wei Lü, Desheng Li, Lang Tan, Peiguang Jing, Yuting Su. Microvideo Multilabel Learning Model Based on Multiview Low-Rank Representation[J]. Laser & Optoelectronics Progress, 2020, 57(22): 221012
    Download Citation