Home | News | Team | Projects | Research | Standards | Demos | Databases | Parteners | Contact      嶄猟
SEARCH        
 
Baseline Results on APiS 1.0 Database

Regarding binary attribute classification, we extract color and texture features based on sliding window and then use Gentle AdaBoost algorithm to train classifiers. Regarding multi-class attribute classification, we extract color features and use a weighted K Nearest Neighbors (KNN) algorithm to accomplish the multi-class classification. The detail please see our paper [1].

1. Binary Attribute Classification Result

Sample

Sample

Figure 1 Average ROC curve of each binary attribute classification. Here, * indicates the corresponding average ROC curve has maximum AUC value.

Sample

Table 2 Average recall rate of each binary attribute when using single type of feature and the fusion feature at the average false positive rate of 0.1.

2. Multi-class Attribute Classification Result

Sample

Figure 2 Average ROC curves of upper-body and lower-body clothing color attribute.

Sample

Figure 3 The curve of average recall rate and false positive rate for each color in upper-body and lower-body clothing color attributes.

Reference
[1] Jianqing Zhu, Shengcai Liao, Zhen Lei, Dong Yi and Stan Z. Li, ^Pedestrian Attribute Classification in Surveillance: Database and Evaluation ̄. In ICCV workshop on Large-Scale Video Search and Mining (LSVSM'13), Sydney, December, 2013.[pdf]

   Introduction
   Iris Databases
   Gait Databases
   HFB Face Databases
   NIR-VIS 2.0 Databases
   NIR Face Databases
   BIT Face Databases
   Fingerprint Databases
   Handwriting Databases
   Action Databases
   Palmprint Databases
   Multi-spectral Palmprint Databases
   APiS 1.0 Databases
Copy rigth All right reserved 2005 Center for Biometrics and Security Research
Center for Biometrics and Security Research 12th Floor,Institute of Automation chinese Academy of Sciences
P.O.Box2728Beijing 100080 P.R.China
Tel:010-62632259 Fax:010-62632259 E-MAIL:hjzhang@cbsr.ia.ac.cn