Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification
Authors:
- Zhongdao WangZWZhongdao Wang
- Luming TangLTLuming Tang
- Xihui LiuXihui Liu
- Zhuliang YaoZYZhuliang Yao
- Shuai YiSYShuai Yi
- Jing ShaoJSJing Shao
- Junjie YanJYJunjie Yan
- Shengjin WangSWShengjin Wang
- Hongsheng LiHLHongsheng Li
- Xiaogang WangXWXiaogang Wang
- 1 October 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 379-387
Abstract
In this paper, we tackle the vehicle Re-identification (ReID) problem which is of great importance in urban surveillance and can be used for multiple applications. In our vehicle ReID framework, an orientation invariant feature embedding module and a spatial-temporal regularization module are proposed. With orientation invariant feature embedding, local region features of different orientations can be extracted based on 20 key point locations and can be well aligned and combined. With spatial-temporal regularization, the log-normal distribution is adopted to model the spatial-temporal constraints and the retrieval results can be refined. Experiments are conducted on public vehicle ReID datasets and our proposed method achieves state-of-the-art performance. Investigations of the proposed framework is conducted, including the landmark regressor and comparisons with attention mechanism. Both the orientation invariant feature embedding and the spatio-temporal regularization achieve considerable improvements.Keywords
This publication has 22 references indexed in Scilit:
- Spindle Net: Person Re-identification with Human Body Region Guided Feature Decomposition and Fusion2017
- Hierarchical Gaussian Descriptor for Person Re-identification2016
- BoxCars: 3D Boxes as CNN Input for Improved Fine-Grained Vehicle Recognition2016
- Scalable Person Re-identification: A Benchmark2015
- Kernelized Saliency-Based Person Re-Identification Through Multiple Metric LearningIEEE Transactions on Image Processing, 2015
- FaceNet: A unified embedding for face recognition and clustering2015
- Going deeper with convolutions2015
- Transferring a semantic representation for person re-identification and search2015
- Face Alignment at 3000 FPS via Regressing Local Binary Features2014
- Large scale metric learning from equivalence constraints2012