Table 2:mAP on CIFAR-10 data set.The best mAP among SBQ.HH and DBQ under the same setting is shown in bold face. bits 32 64 128 256 SBO HH DBO SBO HH DBO SBO HH DBO SBO HH DBO ITo 0.27160.22400.2220 0.3293 0.30060.3350 0.3593 0.3826 0.439明 0.3727 0.4140 0.5221 SH 0.05110.07420.1114 0.06380.09360.1717 0.09980.1209 0.2501 0.1324 0.1697 0.3337 PCA 0.03570.0646 0.1072 0.03111 0.07330.1541 0.0261 0.0835 0.1966 0.0217 0.1127 0.2053 LSH 0.11920.0665 0.0660 0.1882 0.14570.1588 0.2837 0.2601 03153 0.3480 0.3640 0.4680 SIKH 0.04170.0359 0.04660.0953 0.09110.1063 0.1836 0.1969 0.2263 0.3677 0.3601 0.3975 ITQ-SBQ T0-S80 ITQ-SBQ -T0-S80 -TO-HH HH -TO-HH -IT0- +=0-DB0 ITQ-DB -◆-T0-DE 02 02 04 0.8 04 0 (a)ITQ32 bits (b)ITQ 64 bits (c)ITQ 128 bits (d)ITQ 256 bits Figure 2:Precision-recall curve for ITO on 22K LabelMe data set --SH-SBQ -SH-S8Q SH-SBO --SH-SBO ◆S4-HH SH-HH SH-HH SH-HH +-SH-D80 SH-DBQ SH-DBO SH-DBO 02 02 Real时 02 04 08 02 ecal时 04 (a)SH32 bits (b)SH 64 bits (c)SH 128 bits (d)SH 256 bits Figure 3:Precision-recall curve for SH on 22K LabelMe data set -LSH LSH-S80 -LSH- LSH-SBO LSH-Hm LSH-MH LSH-HR SH-HH LSH-DBO -◆-LS+D30 LSH-DBO →-LS+DB0 0.5 02 04 06 0.8 02 05 02 04 05 Recall Recall Recal (a)LSH32 bits (b)LSH 64 bits (c)LSH 128 bits (d)LSH 256 bits Figure 4:Precision-recall curve for LSH on 22K LabelMe data setTable 2: mAP on CIFAR-10 data set. The best mAP among SBQ, HH and DBQ under the same setting is shown in bold face. # bits 32 64 128 256 SBQ HH DBQ SBQ HH DBQ SBQ HH DBQ SBQ HH DBQ ITQ 0.2716 0.2240 0.2220 0.3293 0.3006 0.3350 0.3593 0.3826 0.4395 0.3727 0.4140 0.5221 SH 0.0511 0.0742 0.1114 0.0638 0.0936 0.1717 0.0998 0.1209 0.2501 0.1324 0.1697 0.3337 PCA 0.0357 0.0646 0.1072 0.0311 0.0733 0.1541 0.0261 0.0835 0.1966 0.0217 0.1127 0.2053 LSH 0.1192 0.0665 0.0660 0.1882 0.1457 0.1588 0.2837 0.2601 0.3153 0.3480 0.3640 0.4680 SIKH 0.0417 0.0359 0.0466 0.0953 0.0911 0.1063 0.1836 0.1969 0.2263 0.3677 0.3601 0.3975 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision ITQ−SBQ ITQ−HH ITQ−DBQ (a) ITQ 32 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision ITQ−SBQ ITQ−HH ITQ−DBQ (b) ITQ 64 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision ITQ−SBQ ITQ−HH ITQ−DBQ (c) ITQ 128 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision ITQ−SBQ ITQ−HH ITQ−DBQ (d) ITQ 256 bits Figure 2: Precision-recall curve for ITQ on 22K LabelMe data set 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision SH−SBQ SH−HH SH−DBQ (a) SH 32 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision SH−SBQ SH−HH SH−DBQ (b) SH 64 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision SH−SBQ SH−HH SH−DBQ (c) SH 128 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision SH−SBQ SH−HH SH−DBQ (d) SH 256 bits Figure 3: Precision-recall curve for SH on 22K LabelMe data set 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision LSH−SBQ LSH−HH LSH−DBQ (a) LSH 32 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision LSH−SBQ LSH−HH LSH−DBQ (b) LSH 64 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision LSH−SBQ LSH−HH LSH−DBQ (c) LSH 128 bits 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Recall Precision LSH−SBQ LSH−HH LSH−DBQ (d) LSH 256 bits Figure 4: Precision-recall curve for LSH on 22K LabelMe data set