Gait-based gender recognition using pose information for real time applications

Kastaniotis, Dimitrios; Theodorakopoulos, Ilias; Economou, George; Fotopoulos, S.

doi:10.1109/icdsp.2013.6622766

Cited by 36 publications

(27 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As opposed to the authors in (Kastaniotis et al, 2013) here we evaluate our method on both human gait and gender recognition. Next we provide a short description of the dataset.…”

Section: Resultsmentioning

confidence: 95%

“…In order to evaluate our method we used a publicly available dataset originally presented in (Kastaniotis et al, 2013). As opposed to the authors in (Kastaniotis et al, 2013) here we evaluate our method on both human gait and gender recognition.…”

Section: Resultsmentioning

confidence: 99%

“…In order to efficiently represent the dynamic characteristics of gait we select a number of limbs that are found to be mostly informative (Theodorakopoulos et al, 2014;Kastaniotis et al, 2013). These are the upper and lower arm and upper and lower leg in both sides.…”

Section: Feature Extraction and Representationmentioning

confidence: 99%

“…In (Ball et al, 2012) authors incorporated the mean, standard deviation and maximum value of three angles for each one of the left and right legs, a total of 18 features that were whitened before running K-means. More recently (Kastaniotis et al, 2013) targeting the task of real time gender recognition, incorporated the skeleton data captured from Kinect sensor in order to identify the gender of a walking person by aggregating features into histograms and then classifying samples using an SVM classifier. The framework proposed in this work enhances the initial findings of (Kastaniotis et al, 2013) that "gait encompasses information about gender" and extends these to the field of identity recognition.…”

Section: Related Workmentioning

confidence: 99%

“…In this context and inspired by our prior work on human motion analysis (Kastaniotis et al, 2013;Theodorakopoulos et al, 2014), in this paper we propose a framework for pose-based gait recognition and identification as well as gender recognition. Our framework utilizes a multiple level pose data processing scheme, where pose data acquired using Microsofts" Kinect sensor are used in order to recognize a walking person.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A framework for gait-based recognition using Kinect

Kastaniotis

Theodorakopoulos

Theoharatos

et al. 2015

Pattern Recognition Letters

Self Cite

View full text Add to dashboard Cite

“…As opposed to the authors in (Kastaniotis et al, 2013) here we evaluate our method on both human gait and gender recognition. Next we provide a short description of the dataset.…”

Section: Resultsmentioning

confidence: 95%

Section: Resultsmentioning

confidence: 99%

Section: Feature Extraction and Representationmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A framework for gait-based recognition using Kinect

Kastaniotis

Theodorakopoulos

Theoharatos

et al. 2015

Pattern Recognition Letters

Self Cite

View full text Add to dashboard Cite

Benchmark RGB-D Gait Datasets: A Systematic Review

2019

View full text Add to dashboard Cite

Human motion analysis has proven to be a great source of information for a wide range of applications. Several approaches for a detailed and accurate motion analysis have been proposed in the literature, as well as an almost proportional number of dedicated datasets. The relatively recent arrival of depth sensors contributed to an increasing interest in this research area and also to the emergence of a new type of motion datasets. This work focuses on a systematic review of publicly available depth-based datasets, encompassing human gait data which is used for person recognition and/or classification purposes. We have conducted this systematic review using the Scopus database. The herein presented survey, which to the best of our knowledge is the first one dedicated to this type of datasets, is intended to inform and aid researchers on the selection of the most suitable datasets to develop, test and compare their algorithms.

show abstract

Gender Recognition from 3D Shape Parameters

Martinelli

Garau

Conci

2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Gender recognition from images is generally approached by extracting the salient visual features of the observed subject, either focusing on the facial appearance or by analyzing the full body. In real-world scenarios, image-based gender recognition approaches tend to fail, providing unreliable results. Facebased methods are compromised by environmental conditions, occlusions (presence of glasses, masks, hair), and poor resolution. Using a full-body perspective leads to other downsides: clothing and hairstyle may not be discriminative enough for classification, and background cluttering could be problematic. We propose a novel approach for body-shape-based gender classification. Our contribution consists in introducing the so-called Skinned Multi-Person Linear model (SMPL) as 3D human mesh. The proposed solution is robust to poor image resolution and the number of features for the classification is limited, making the recognition task computationally affordable, especially in the classification stage, where less complex learning architectures can be easily trained. The obtained information is fed to an SVM classifier, trained and tested using three different datasets, namely (i) FVG, containing videos of walking subjects, (ii) AMASS, collected by converting MOCAP data of people performing different activities into realistic 3D human meshes, and (iii) SURREAL, characterized by synthetic human body models. Additionally, we demonstrate that our approach leads to reliable results even when the parametric 3D mesh is extracted from a single image. Considering the lack of benchmarks in this area, we trained and tested the FVG dataset with a pre-trained Resnet50, for comparing our model-based method with an imagebased approach.

show abstract

Gait-based gender recognition using pose information for real time applications

Cited by 36 publications

References 14 publications

A framework for gait-based recognition using Kinect

A framework for gait-based recognition using Kinect

Benchmark RGB-D Gait Datasets: A Systematic Review

Gender Recognition from 3D Shape Parameters

Contact Info

Product

Resources

About