Multi-Channel Pyramid Person Matching Network for Person Re-Identification

Authors: Chaojie Mao, Yingming Li, Yaqing Zhang, Zhongfei Zhang, Xi Li

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on several benchmark datasets demonstrate the effectiveness of our approach against the state-of-the-art literature, especially on the rank-1 recognition rate.
Researcher Affiliation Collaboration 1 College of Information Science & Electronic Engineering, Zhejiang University, Hangzhou, China 2 College of Computer Science and Technology, Zhejiang University, Hangzhou, China 3 Alibaba-Zhejiang University Joint Institute of Frontier Technologies, Hangzhou, China
Pseudocode No The paper includes network architecture diagrams (e.g., Figure 2) and descriptive text, but no formally structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement about releasing source code or a link to a code repository.
Open Datasets Yes We evaluate the proposed architecture and compare our results with those of the state-of-the-art approaches on six person Re-ID datasets, namely CUHK03 (Li et al. 2014), CUHK01 (Li, Zhao, and Wang 2012), VIPe R (Gray and Tao 2008), PRID450s (Roth et al. 2014), i-LIDS (Office 2008) and PRID2011 (Hirzer et al. 2011). ... Table lists the description of each dataset and our experimental settings with the training and testing splits.
Dataset Splits Yes Table lists the description of each dataset and our experimental settings with the training and testing splits. ... For the CUHK01 dataset, we report results on two different settings: 100 test IDs, and 486 test IDs.
Hardware Specification Yes The proposed architecture is implemented on the widely used deep learning framework Caffe (Jia et al. 2014) with an NVIDIA TITAN X GPU.
Software Dependencies No The paper mentions using "Caffe (Jia et al. 2014)" but does not provide specific version numbers for Caffe or any other software dependencies.
Experiment Setup Yes The parameters for training SC-PPMN, CTM-PPMN and MCPPMN are listed in Table 2. ... Table 2: The parameters for training. Parameters: Maximum Iteration, Batch Size, Momentum, Weight Decay, Base Learning Rate.