Skip to Main content Skip to Navigation

Image-based Gender Estimation from Body and Face across Distances

Abstract : Gender estimation has received increased attention due to its use in a number of pertinent security and commercial applications. Automated gender estimation algorithms are mainly based on extracting representative features from face images. In this work we study gender estimation based on information deduced jointly from face and body, extracted from single-shot images. The approach addresses challenging settings such as low-resolution-images, as well as settings when faces are occluded. Specifically the face-based features include local binary patterns (LBP) and scale-invariant feature transform (SIFT) features, projected into a PCA space. The features of the novel body-based algorithm proposed in this work include continuous shape information extracted from body silhouettes and texture information retained by HOG descriptors. Support Vector Machines (SVMs) are used for classification for body and face features. We conduct experiments on images extracted from video-sequences of the Multi-Biometric Tunnel database, emphasizing on three distance-settings: close, medium and far, ranging from full body exposure (far setting) to head and shoulders exposure (close setting). The experiments suggest that while face-based gender estimation performs best in the close-distance-setting, body-based gender estimation performs best when a large part of the body is visible. Finally we present two score-level-fusion schemes of face and body-based features, outperforming the two individual modalities in most cases.
Complete list of metadatas

Cited literature [39 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01384324
Contributor : Antitza Dantcheva <>
Submitted on : Wednesday, October 19, 2016 - 4:41:54 PM
Last modification on : Thursday, November 14, 2019 - 5:38:06 PM

File

bare_conf_v6_camera_ready.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01384324, version 1

Collections

Citation

Ester Gonzalez-Sosa, Antitza Dantcheva, Ruben Vera-Rodriguez, Jean-Luc Dugelay, François Brémond, et al.. Image-based Gender Estimation from Body and Face across Distances. 23rd International Conference on Pattern Recognition (ICPR 2016): "Image analysis and machine learning for scene understanding", Dec 2016, Cancun, Mexico. ⟨hal-01384324⟩

Share

Metrics

Record views

362

Files downloads

1047