Person Re-identification Using Convolutional Neural Networks

Download
2021-9-6
Aksu, Fatih
Person re-identification is an important computer vision topic particularly for surveillance system applications. The main aim of person re-identification is to identify previously observed individuals over a camera network with nonoverlapping occurrences. Various approaches have been introduced to overcome this problem. This study has two distinct key contributions. First, a new part-based method for person re-identification, which combines semantically partitioned body part masks with a convolutional neural network, is proposed. With this study, it is observed that the proposed body part-based system has promising results, and has advantages over the baseline person re-identification systems. The second contribution of this study is to investigate and compare potential lightweight convolutional neural networks suitable for person re-identification task. Results show that some lightweight convolutional neural networks can be used instead of more generalized deeper networks. Well-designed lightweight convolutional neural networks may have higher accuracy with lower computational cost for person reidentification.

Suggestions

Quality Enhancement of Computed Tomography Images of Porous Media Using Convolutional Neural Networks
Yıldırım, Ertuğrul Umut; Uğur, Ömür; Glatz, Guenther; Department of Scientific Computing (2022-2-11)
Computed tomography has been widely used in clinical and industrial applications as a non-destructive visualization technology. The quality of computed tomography scans has a strong effect on the accuracy of the estimated physical properties of the investigated sample. X-ray exposure time is a crucial factor for scan quality. Ideally, long exposure time scans, yielding large signal-to-noise ratios, are available if physical properties are to be delineated. However, especially in micro-computed tomography ap...
UTILIZATION OF SPATIAL INFORMATION FOR POINT CLOUD SEGMENTATION
Akman, Oytun; Bayramoglu, Neslihan; Alatan, Abdullah Aydın; Jonker, Pieter (2010-06-09)
Object segmentation has an important role in the field of computer vision for semantic information inference. Many applications such as 3DTV archive systems, 3D/2D model fitting, object recognition and shape retrieval are strongly dependent to the performance of the segmentation process. In this paper we present a new algorithm for object localization and segmentation based on the spatial information obtained via a Time-of-Flight (TOF) camera. 3D points obtained via a TOF camera are projected onto the major...
Face Recognition Based on Embedding Learning
Karaman, Kaan; Koc, Aykut; Alatan, Abdullah Aydın (2018-09-11)
Face recognition is a key task of computer vision research that has been employed in various security and surveillance applications. Recently, the importance of this task has risen with the improvements in the quality of sensors of cameras, as well as with the increasing coverage of camera networks setup everywhere in the cities. Moreover, biometry-based technologies have been developed for the last three decades and have been available on many devices such as the mobile phones. The goal is to identify peop...
An FPGA based high performance optical flow hardware design for computer vision applications
Gultekin, Gokhan Koray; Saranlı, Afşar (2013-05-01)
Optical Flow (OF) information is used in higher level vision tasks in a variety of computer vision applications. However, its use in resource constrained applications such as small-scale mobile robotic platforms is limited because of the high computational complexity involved. The inability to compute the OF vector field in real-time is the main drawback which prevents these applications to efficiently utilize some successful techniques from the computer vision literature. In this work, we present the desig...
Comparison of feature sets using multimedia translation
Duygulu, P; Ozcanli, OC; Papernick, N (2003-01-01)
Feature selection is very important for many computer vision applications. However, it is hard to find a good measure for the comparison. In this study, feature sets are compared using the translation model of object recognition which is motivated by the availablity of large annotated data sets. Image regions are linked to words using a model which is inspired by machine translation. Word prediction performance is used to evaluate large numbers of images.
Citation Formats
F. Aksu, “Person Re-identification Using Convolutional Neural Networks,” M.S. - Master of Science, Middle East Technical University, 2021.