Cross-Modality Person Re-Identification with Generative Adversarial Training
Authors: Pingyang Dai, Rongrong Ji, Haibin Wang, Qiong Wu, Yuyu Huang
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We have quantized the performance of our work in the newly-released SYSU RGB-IR Re-ID benchmark, and have reported superior performance, i.e., Cumulative Match Characteristic curve (CMC) and Mean Average Precision (MAP), over the state-of-the-art works [Wu et al., 2017], at least 12.17% and 11.85% respectively. |
| Researcher Affiliation | Academia | 1 Fujian Key Laboratory of Sensing and Computing for Smart City, Xiamen University, China 2 School of Information Science and Engineering, Xiamen University, China {pydai, rrji}@xmu.edu.cn, {haibin, qiong, huangyuyu}@stu.xmu.edu.cn |
| Pseudocode | Yes | Algorithm 1 The Learning of the Proposed cm GAN model |
| Open Source Code | No | The paper does not contain any statement about making its source code available or provide a link to a code repository for the methodology described. |
| Open Datasets | Yes | The SYSU RGB-IR Re-ID dataset1 is the first benchmark for cross-modality (RGB-IR) Re-ID...1http://isee.sysu.edu.cn/project/RGBIRRe ID.htm |
| Dataset Splits | No | The dataset is separated into the training set and the test set, where images of the same person can only appear in either set. And the training set consists of total 32,451 images including 19,659 RGB images and 12,792 IR images. No specific validation set split percentages or counts are mentioned. |
| Hardware Specification | Yes | We use NVIDIA Ge Force 1080Ti graphics cards for our entire experiments. |
| Software Dependencies | No | The paper mentions a 'standard deep neural network framework' and specifies a GPU model, but does not provide specific software names with version numbers for libraries, frameworks, or languages used in the experiments. |
| Experiment Setup | Yes | The batch size is set to 20. Empirically after testing several groups of parameter combinations, the generative model training step K is set to be 5. The adaptive parameter γ controls the weight of the discriminator loss and is fixed to be 0.05. We set margin ξ in cross-modality triplet loss in the range [0.7, 0.9, 1.2, 1.4, 1.5, 1.6]... We get the best result when generator lr is set to be 0.0001 with discriminator lr is set to be 0.001 respectively. Therefore, we set the training epoch to 2000 and more than it. Empirically, we set α and β by 1:1 proportion... |