• Acta Optica Sinica
  • Vol. 39, Issue 7, 0728001 (2019)
Chen Wu1, Hongwei Wang2, Zhiqiang Wang2, Yuwei Yuan3..., Yu Liu2, Hong Cheng2 and Jicheng Quan2,*|Show fewer author(s)
Author Affiliations
  • 1 University of Naval Aviation, Yantai, Shandong 264000, China
  • 2 Aviation University of Air Force, Changchun, Jilin 130022, China
  • 3 The 91977 of Peoples Liberation Army of China, Beijing 102200, China
  • show less
    DOI: 10.3788/AOS201939.0728001 Cite this Article Set citation alerts
    Chen Wu, Hongwei Wang, Zhiqiang Wang, Yuwei Yuan, Yu Liu, Hong Cheng, Jicheng Quan. Zero-Shot Classification for Remote Sensing Scenes Based on Locality Preservation[J]. Acta Optica Sinica, 2019, 39(7): 0728001 Copy Citation Text show less
    Whole framework of the proposed algorithm
    Fig. 1. Whole framework of the proposed algorithm
    Images of several UCM classes. (a) Farmland; (b) airplane; (c) baseball diamond; (d) dense residential; (e) freeway; (f) harbor; (g) storage tanks; (h) tennis court; (i) overpass; (j) golf course
    Fig. 2. Images of several UCM classes. (a) Farmland; (b) airplane; (c) baseball diamond; (d) dense residential; (e) freeway; (f) harbor; (g) storage tanks; (h) tennis court; (i) overpass; (j) golf course
    Images of RSSCN7 classes. (a) Grass; (b) river lake; (c) industry; (d) field; (e) forest; (f) residential; (g) parking
    Fig. 3. Images of RSSCN7 classes. (a) Grass; (b) river lake; (c) industry; (d) field; (e) forest; (f) residential; (g) parking
    The t-SNE graphs of Xu and different Zu on GoogLeNet feature. (a) Xu; (b) Zu+I(λ=0, μ=0); (c) Zu+I(λ=1, μ=0); (d) Zu+T(λ=0, μ=1); (e) Zu+T(λ=1, μ=1)
    Fig. 4. The t-SNE graphs of Xu and different Zu on GoogLeNet feature. (a) Xu; (b) Zu+I(λ=0, μ=0); (c) Zu+I(λ=1, μ=0); (d) Zu+T(λ=0, μ=1); (e) Zu+T(λ=1, μ=1)
    The t-SNE graphs of Xu and different Zu on VGGNet feature. (a) Xu; (b) Zu+I(λ=0, μ=0); (c) Zu+I(λ=1, μ=0); (d) Zu+T(λ=0, μ=1); (e) Zu+T(λ=1, μ=1)
    Fig. 5. The t-SNE graphs of Xu and different Zu on VGGNet feature. (a) Xu; (b) Zu+I(λ=0, μ=0); (c) Zu+I(λ=1, μ=0); (d) Zu+T(λ=0, μ=1); (e) Zu+T(λ=1, μ=1)
    OA values of proposed algorithm under different λ and μ
    Fig. 6. OA values of proposed algorithm under different λ and μ
    The ζsum/OA curves of algorithm under GoogLeNet feature
    Fig. 7. The ζsum/OA curves of algorithm under GoogLeNet feature
    The ζsum/OA curves of algorithm under VGGNet feature
    Fig. 8. The ζsum/OA curves of algorithm under VGGNet feature
    Accuracies of different classes of RSSCN7 produced by I type ZSC methods. (a) GoogLeNet; (b) VGGNet
    Fig. 9. Accuracies of different classes of RSSCN7 produced by I type ZSC methods. (a) GoogLeNet; (b) VGGNet
    Accuracies of different classes of RSSCN7 produced by T type ZSC methods.(a) GoogLeNet; (b)VGGNet
    Fig. 10. Accuracies of different classes of RSSCN7 produced by T type ZSC methods.(a) GoogLeNet; (b)VGGNet
    MethodI/TOA /%
    GoogLeNetVGGNet
    LatEmI18.7914.15
    BiDiLELI14.5114.47
    JLSEI36.6133.45
    SSEI39.8637.36
    DMaPI34.0032.07
    SAEI34.2035.60
    RKTI35.4732.07
    RSZSC_II43.4646.50
    UDAT42.6143.59
    TMET41.9642.44
    SMST37.1539.18
    RSZSC_TT50.6753.29
    Table 1. OA values of ZSC methods on RSSCN7
    MethodI/TTime /s
    LatEmI17.71
    BiDiLELI22.59
    JLSEI67.49
    SSEI14.35
    DMaPI396.45
    SAEI16.21
    RKTI22.53
    RSZSC_II11.68
    UDAT56.37
    TMET76.91
    SMST49.28
    RSZSC_TT14.70
    Table 2. Computing time of different ZSC algorithms on GoogLeNet feature
    Chen Wu, Hongwei Wang, Zhiqiang Wang, Yuwei Yuan, Yu Liu, Hong Cheng, Jicheng Quan. Zero-Shot Classification for Remote Sensing Scenes Based on Locality Preservation[J]. Acta Optica Sinica, 2019, 39(7): 0728001
    Download Citation