Skip to main content

Advertisement

Table 1 The rank-1 precious (%) and mAP (%) on the Market-1501 database

From: Learning deep features from body and parts for person re-identification in camera networks

Method Single Query Multi Query
  rank-1 mAP rank-1 mAP
DADM [26] 39.40 19.6 49.00 25.8
BoW+KISSME [19] 44.42 20.26
MST-CNN [27] 45.1 55.40
MR-CNN [28] 45.58 26.11 56.59 32.26
FisherNet [29] 48.15 29.94
CAN [30] 48.24 24.43
SL [31] 51.90 26.11 56.59 32.26
SCSP [31] 51.90 26.35
DNS [32] 55.43 29.87 71.56 46.03
S-LSTM [33] 61.60 35.3
Gate Reid [34] 65.88 39.55 76.04 48.45
SOMAnet [35] 73.87 47.89 81.29 56.98
MSCAN [36] 75.45 52.41 83.43 62.03
PIE [37] 78.65 53.87
Verif.-Identif [16] 79.51 59.87 85.84 70.33
Part-based features 36.25 14.47 37.86 13.51
Body-based features 80.29 59.34 86.81 71.18
DFBP 81.71 60.86 87.02 72.21
  1. The data in italics are the best result in each evaluation protocol