Method RGB-SOBS on the SBM-RGBD dataset
Contact name: Lucia Maddalena
Contact email: lucia.maddalena@cnr.it
Contact Institution: National Research Council of Italy
Method processing time: ~4 fps for a 640x480 video (including I/O) with C++ code running on a Intel Core i5 2.7GHz
Note: Uses only RGB data
Reference: L. Maddalena, A. Petrosino, The SOBS algorithm: What are the limits?,
2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp.21-26, 16-21 June 2012.
Get all result files:
here.
A) Sequence-by-sequence results
Name | Recall | Specificity | FPR | FNR | PWC | Precision | F-Measure |
Bootstrapping |
BootStrapping_ds | 0.9155 | 0.9942 | 0.0058 | 0.0845 | 1.3206 | 0.9423 | 0.9287 |
adl24cam0 | 0.7030 | 0.9917 | 0.0083 | 0.2970 | 2.5818 | 0.8454 | 0.7676 |
bear_front | 0.6026 | 0.9588 | 0.0412 | 0.3974 | 13.4976 | 0.8396 | 0.7016 |
fall01cam0 | 0.8935 | 0.9807 | 0.0193 | 0.1065 | 2.4133 | 0.7316 | 0.8045 |
fall20cam0 | 0.8969 | 0.9814 | 0.0186 | 0.1031 | 2.2970 | 0.7234 | 0.8008 |
ColorCamouflage |
Cespatx_ds | 0.8111 | 0.9997 | 0.0003 | 0.1889 | 0.4202 | 0.9797 | 0.8875 |
Hallway | 0.5782 | 0.9629 | 0.0371 | 0.4218 | 5.7161 | 0.4624 | 0.5138 |
colorCam1 | 0.1045 | 0.9997 | 0.0003 | 0.8955 | 20.0782 | 0.9906 | 0.1891 |
colorCam2 | 0.2304 | 0.9443 | 0.0557 | 0.7696 | 37.9469 | 0.7745 | 0.3551 |
DepthCamouflage |
DCamSeq1 | 0.9969 | 0.9770 | 0.0230 | 0.0031 | 2.1855 | 0.7197 | 0.8359 |
DCamSeq2 | 0.9942 | 0.9912 | 0.0088 | 0.0058 | 0.8676 | 0.8378 | 0.9093 |
Despatx_ds | 0.9001 | 0.9985 | 0.0015 | 0.0999 | 1.0473 | 0.9831 | 0.9397 |
Wall | 0.9990 | 0.9757 | 0.0243 | 0.0010 | 2.2233 | 0.8011 | 0.8892 |
IlluminationChanges |
ChairBox | 0.7684 | 0.9986 | 0.0014 | 0.2316 | 2.6509 | 0.9848 | 0.8633 |
Ls_ds | 0.0000 | 0.9212 | 0.0788 | 0.0000 | 7.8775 | 0.0000 | 0.0000 |
TimeOfDay_ds | 0.0000 | 0.9789 | 0.0211 | 0.0000 | 2.1150 | 0.0000 | 0.0000 |
genSeq1 | 0.9781 | 0.9875 | 0.0125 | 0.0219 | 1.3655 | 0.9188 | 0.9475 |
IntermittentMotion |
Shelves | 0.9023 | 0.9294 | 0.0706 | 0.0977 | 7.1144 | 0.2033 | 0.3319 |
Sleeping_ds | 0.9020 | 0.9339 | 0.0661 | 0.0980 | 7.2499 | 0.7744 | 0.8334 |
abandoned1 | 0.9785 | 0.9778 | 0.0222 | 0.0215 | 2.2179 | 0.5309 | 0.6883 |
abandoned2 | 0.9909 | 0.7622 | 0.2378 | 0.0091 | 21.0108 | 0.3650 | 0.5335 |
movedBackground1 | 0.9510 | 0.9640 | 0.0360 | 0.0490 | 3.6208 | 0.2838 | 0.4371 |
movedBackground2 | 0.8344 | 0.8499 | 0.1501 | 0.1656 | 15.1125 | 0.2753 | 0.4140 |
OutOfRange |
MultiPeople1 | 0.9586 | 0.9702 | 0.0298 | 0.0414 | 3.0693 | 0.7321 | 0.8302 |
MultiPeople2 | 0.9832 | 0.9945 | 0.0055 | 0.0168 | 0.6429 | 0.9407 | 0.9615 |
TopViewLab1 | 0.7687 | 0.9943 | 0.0057 | 0.2313 | 1.3208 | 0.8229 | 0.7949 |
TopViewLab2 | 0.9067 | 0.9964 | 0.0036 | 0.0933 | 0.7012 | 0.9107 | 0.9087 |
TopViewLab3 | 0.8339 | 0.9927 | 0.0073 | 0.1661 | 1.0710 | 0.7120 | 0.7681 |
Shadows |
Shadows_ds | 0.7565 | 0.9963 | 0.0037 | 0.2435 | 1.8413 | 0.9311 | 0.8347 |
fall01cam1 | 0.9789 | 0.9966 | 0.0034 | 0.0211 | 0.4480 | 0.9513 | 0.9649 |
genSeq2 | 0.9768 | 0.9903 | 0.0097 | 0.0232 | 1.1179 | 0.9269 | 0.9512 |
shadows1 | 0.9723 | 0.9794 | 0.0206 | 0.0277 | 2.1610 | 0.8849 | 0.9266 |
shadows2 | 0.9952 | 0.9776 | 0.0224 | 0.0048 | 1.9960 | 0.8759 | 0.9318 |
B) Average results for each category
Name | Recall | Specificity | FPR | FNR | PWC | Precision | F-Measure |
Bootstrapping | 0.8023 | 0.9814 | 0.0186 | 0.1977 | 4.4221 | 0.8165 | 0.8007 |
ColorCamouflage | 0.4310 | 0.9767 | 0.0233 | 0.5690 | 16.0404 | 0.8018 | 0.4864 |
DepthCamouflage | 0.9725 | 0.9856 | 0.0144 | 0.0275 | 1.5809 | 0.8354 | 0.8935 |
IlluminationChanges | 0.4366 | 0.9715 | 0.0285 | 0.0634 | 3.5022 | 0.4759 | 0.4527 |
IntermittentMotion | 0.9265 | 0.9028 | 0.0972 | 0.0735 | 9.3877 | 0.4054 | 0.5397 |
OutOfRange | 0.8902 | 0.9896 | 0.0104 | 0.1098 | 1.3610 | 0.8237 | 0.8527 |
Shadows | 0.9359 | 0.9881 | 0.0119 | 0.0641 | 1.5128 | 0.9140 | 0.9218 |
AVG. | 0.7707 | 0.9708 | 0.0292 | 0.1578 | 5.4010 | 0.7247 | 0.7068 |