Divide and Fuse: A Re-ranking Approach for Person Re-identification
Divide and Fuse: A Re-ranking Approach for Person Re-identification
复制标题
DOI：
10.5244/c.31.135
复制
发表时间：
2017-08
期刊：
影响因子：
--
通讯作者：
Rui Yu;Zhichao Zhou;S. Bai;X. Bai
中科院分区：
文献类型：
--
作者：
Rui Yu;Zhichao Zhou;S. Bai;X. Bai
由于重新排序是在大规模数据集上提高行人重新识别（re-ID）性能的必要过程，因此特征的多样性对于行人重新识别变得至关重要，因为它对于设计行人描述和基于特征融合的重新排序都很重要。然而，在许多情况下，只有一种类型的行人特征可用。在本文中，我们提出了一种用于人员重识别的“划分和使用”重排序框架。它利用高维特征向量不同部分的多样性进行基于融合的重新排序，而无法访问其他特征。具体来说，给定一幅图像，将提取的特征划分为子特征。然后每个子特征的上下文信息被迭代地编码成新的特征。最后，来自同一图像的新特征被融合到一个向量中以进行重新排序。两个人员重新识别基准的实验结果证明了所提出框架的有效性。特别是，我们的方法优于 Market-1501 数据集上最先进的方法。
As re-ranking is a necessary procedure to boost person re-identification (re-ID) performance on large-scale datasets, the diversity of feature becomes crucial to person reID for its importance both on designing pedestrian descriptions and re-ranking based on feature fusion. However, in many circumstances, only one type of pedestrian feature is available. In this paper, we propose a "Divide and use" re-ranking framework for person re-ID. It exploits the diversity from different parts of a high-dimensional feature vector for fusion-based re-ranking, while no other features are accessible. Specifically, given an image, the extracted feature is divided into sub-features. Then the contextual information of each sub-feature is iteratively encoded into a new feature. Finally, the new features from the same image are fused into one vector for re-ranking. Experimental results on two person re-ID benchmarks demonstrate the effectiveness of the proposed framework. Especially, our method outperforms the state-of-the-art on the Market-1501 dataset.