Robust Depth-Based Person Re-Identification

Abstract

Abstract-Person re-identification (re-id) aims to match people across non-overlapping camera views. So far the RGB-based appearance is widely used in most existing works. However, when people appeared in extreme illumination or changed clothes, the RGB appearance-based re-id methods tended to fail. To overcome this problem, we propose to exploit depth information to provide more invariant body shape and skeleton information regardless of illumination and color change. More specifically, we exploit depth voxel covariance descriptor and further propose a locally rotation invariant depth shape descriptor called Eigen-depth feature to describe pedestrian body shape. We prove that the distance between any two covariance matrices on the Riemannian manifold is equivalent to the Euclidean distance between the corresponding Eigen-depth features. Furthermore, we propose a kernelized implicit feature transfer scheme to estimate Eigendepth feature implicitly from RGB image when depth information is not available. We find that combining the estimated depth features with RGB-based appearance features can sometimes help to better reduce visual ambiguities of appearance features caused by illumination and similar clothes. The effectiveness of our models was validated on publicly available depth pedestrian datasets as compared to related methods for person re-identification.

HSME: Hypersphere Manifold Embedding for Visible Thermal Person Re-Identification

“…Nguyen et al (Nguyen et al 2017) firstly applied person re-identification models to visiblethermal images. Wu et al (Wu, Zheng, and Lai 2017) designed a depth shape descriptor which is robust to rotation and noises. Meanwhile, Lin et al (Lin et al 2017) combined attribute information with image information for visible images.…”

Attention in Multimodal Neural Networks for Person Re-identification

“…Multimodal fusion of RGB and depth information is rarely considered in person re-id [18,21,27]. Liciotti et al [18] propose a combination of hand-crafted RGB and depth features to capture both color, texture and anthropometric information.…”

“…Liciotti et al [18] propose a combination of hand-crafted RGB and depth features to capture both color, texture and anthropometric information. RGB-D based hand-crafted features are also proposed by Wu et al [27] who extract a rotation invariant Eigen-depth feature and fuse it with low-level color and texture features [17]. Only two previous proposals fuse RGB and depth features using a CNN [10,13].…”

Hybrid framework for detection of human face based on haar-like feature

“…There are over in excess of 150 methodologies for face detection. These methodologies are characterized into four distinct classifications, they are Knowledge-based, Feature invariant, Template matching, and Appearance-based [2]. In the accompanying, a short audit of these four classifications is given.…”