Research | Human Detection Method | Precision | Accuracy | Throughput | Comments |
Our method | faster_rcnn_resnet101_v1 KCF/HOG | 95.4 | 92.17% | 16 | Combined adaptive detection/tracking |
Lestari P. [9] | Viola Jones Algorithm and Chan-Vese active contour detection | 99.2 | 95.3 | 0.18 | - |
Xia et al. [32] | Depth information combined with 2-D head contour model and a 3-D head surface model | 93.5 | 88.4 | 0.11 | Depth information is taken from a suitable camera |
Choi et al. [33] | Parameterized heuristics filtering/HOD/SVM | 92.5 | 82.3 | 0.16 | - |
Ikemura et al. [34] | Relational Depth Similarity Features (RDSF) | 90.4 | 87.5 | 10 | The depth information is obtained from a TOF camera |
Chowdhury et al. [7] | Background model/ HSI color model and color correlogram/ImHOG | 93.5 | - | 2.17 | - |
Lee, et al. [8] | R-CNN | 92.9 | - | 8 | - |