Group Cost-Sensitive Boosting for Multi-Resolution Pedestrian Detection

Authors: Chao Zhu, Yuxin Peng

AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The proposed approach is evaluated on the challenging Caltech pedestrian benchmark, and outperforms other state-of-the-art on different resolution-specific test sets.
Researcher Affiliation Academia Chao Zhu and Yuxin Peng Institute of Computer Science and Technology, Peking University Beijing 100871, China {zhuchao, pengyuxin}@pku.edu.cn
Pseudocode Yes Algorithm 1 Group Cost-Sensitive Ada Boost
Open Source Code No The paper mentions evaluation software but does not provide access to the source code for the methodology described in this paper.
Open Datasets Yes The experiments are conducted on the Caltech pedestrian detection benchmark (Dollár et al. 2012)
Dataset Splits Yes The pedestrian detector is trained on the training set (set00-set05), and the detection results are reported on the test set (set06-set10). ... The optimal value of the costs for different groups are selected from Cfp = 1, Cfnh [1 : 0.1 : 10] and Cfnl [Cfnh : 0.1 : Cfnh + 10] by cross-validation.
Hardware Specification No The paper does not provide specific hardware details (like CPU/GPU models or memory amounts) used for running its experiments.
Software Dependencies No The paper mentions using LDCF detector and Ada Boost but does not provide specific version numbers for software dependencies.
Experiment Setup Yes The optimal value of the costs for different groups are selected from Cfp = 1, Cfnh [1 : 0.1 : 10] and Cfnl [Cfnh : 0.1 : Cfnh + 10] by cross-validation. 4096 weak classifiers are trained and combined to a strong classifier... Three bootstrapping stages are applied with 25,000 additional hard negatives each time.