A deep person re-identification model with multi visual-semantic information embedding
- PDF / 3,258,978 Bytes
- 18 Pages / 439.37 x 666.142 pts Page_size
- 56 Downloads / 194 Views
A deep person re-identification model with multi visual-semantic information embedding Xiaopei Wang 1 & Xiaoxia Liu 1 & Jun Guo 1 Yun Xiao 1 & Baoying Liu 2
1
1
& Jiaxiang Zheng & Pengfei Xu &
Received: 29 April 2020 / Revised: 25 August 2020 / Accepted: 17 September 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
The local features of different body parts have been widely used to learn more discriminative representation for person re-identification, which act as either extra visual semantic information or auxiliary means to deal with the issue of misalignment and background bias. However, the existing person re-identification works mainly focuses on the common impact of multiple body parts while failing to explicitly explore the influence of body edge contour. As the edge contour is one of the most significant visual-semantic clues for object detection and person identification in the blurred scene, this paper intentionally explores the effect of edge contour clues on person re-identification and proposes a deep learning framework with multi visual-semantic information embedding, including body parts and edge contour. Meanwhile, we conceive a practical strategy which can effectively fuse the different body part features and reduce the dimensionality of features. Extensive experimental results on four benchmark data sets show that our model has achieved competitive accuracy compared to the state-of-the-art models. Keywords Person re-identification . Edge contour clue . Multiple visual-semantic information embedding . Convolution neural networks
1 Introduction As an essential role in the application of video surveillance and intelligent security, person re-identification (ReID) has drawn more and more attention in recent years. Person ReID aims to automatically identify a person of interest in a large gallery image database, when providing a probe image of person. Different from object detection in static images [17, 33, 36], person ReID faces with various difficulties including
* Jun Guo [email protected] Extended author information available on the last page of the article
Multimedia Tools and Applications
misalignment, occlusion and low resolution. Besides, human pose, cluttered background and illumination change among different camera views can also bring significant difficulties in extracting distinctive attributes. Therefore, an effective person ReID system is obliged to learn representations that are identity specific, context invariant and agnostic with respect to the camera point of view. The past years had witnessed several handcrafted algorithms to learn image representations. However, these algorithms are time consuming for the large data sets and the learned representations are not very discriminative for person ReID. Thanks to the emergence of deep learning and in particular deep convolution neural networks (CNN), significant progress has been achieved for person ReID in recent years. The early CNN models for representations learning usually extract global
Data Loading...
