Vaš brskalnik ne omogoča JavaScript!
JavaScript je nujen za pravilno delovanje teh spletnih strani. Omogočite JavaScript ali pa uporabite sodobnejši brskalnik.
Nacionalni portal odprte znanosti
Odprta znanost
DiKUL
slv
|
eng
Iskanje
Brskanje
Novo v RUL
Kaj je RUL
V številkah
Pomoč
Prijava
Generating bimodal privacy-preserving data for face recognition
ID
Tomašević, Darian
(
Avtor
),
ID
Boutros, Fadi
(
Avtor
),
ID
Damer, Naser
(
Avtor
),
ID
Peer, Peter
(
Avtor
),
ID
Štruc, Vitomir
(
Avtor
)
PDF - Predstavitvena datoteka,
prenos
(3,70 MB)
MD5: E97D16A87A993C8CB4D62862FB68848E
URL - Izvorni URL, za dostop obiščite
https://www.sciencedirect.com/science/article/pii/S0952197624006535
Galerija slik
Izvleček
The performance of state-of-the-art face recognition systems depends crucially on the availability of large-scale training datasets. However, increasing privacy concerns nowadays accompany the collection and distribution of biometric data, which has already resulted in the retraction of valuable face recognition datasets. The use of synthetic data represents a potential solution, however, the generation of privacy-preserving facial images useful for training recognition models is still an open problem. Generative methods also remain bound to the visible spectrum, despite the benefits that multispectral data can provide. To address these issues, we present a novel identity-conditioned generative framework capable of producing large-scale recognition datasets of visible and near-infrared privacy-preserving face images. The framework relies on a novel identity-conditioned dual-branch style-based generative adversarial network to enable the synthesis of aligned high-quality samples of identities determined by features of a pretrained recognition model. In addition, the framework incorporates a novel filter to prevent samples of privacy-breaching identities from reaching the generated datasets and improve both identity separability and intra-identity diversity. Extensive experiments on six publicly available datasets reveal that our framework achieves competitive synthesis capabilities while preserving the privacy of real-world subjects. The synthesized datasets also facilitate training more powerful recognition models than datasets generated by competing methods or even small-scale real-world datasets. Employing both visible and near-infrared data for training also results in higher recognition accuracy on real-world visible spectrum benchmarks. Therefore, training with multispectral data could potentially improve existing recognition systems that utilize only the visible spectrum, without the need for additional sensors.
Jezik:
Angleški jezik
Ključne besede:
image synthesis
,
face-based biometrics
,
privacy-preserving data
,
multispectral recognition
,
generative adversarial networks
Vrsta gradiva:
Članek v reviji
Tipologija:
1.01 - Izvirni znanstveni članek
Organizacija:
FRI - Fakulteta za računalništvo in informatiko
FE - Fakulteta za elektrotehniko
Status publikacije:
Objavljeno
Različica publikacije:
Objavljena publikacija
Leto izida:
2024
Št. strani:
25 str.
Številčenje:
Vol. 133, pt. E, art. 108495
PID:
20.500.12556/RUL-156192
UDK:
004.93:57.087.1
ISSN pri članku:
0952-1976
DOI:
10.1016/j.engappai.2024.108495
COBISS.SI-ID:
194774275
Datum objave v RUL:
13.05.2024
Število ogledov:
297
Število prenosov:
92
Metapodatki:
Citiraj gradivo
Navadno besedilo
BibTeX
EndNote XML
EndNote/Refer
RIS
ABNT
ACM Ref
AMA
APA
Chicago 17th Author-Date
Harvard
IEEE
ISO 690
MLA
Vancouver
:
Kopiraj citat
Objavi na:
Gradivo je del revije
Naslov:
Engineering applications of artificial intelligence
Skrajšan naslov:
Eng. appl. artif. intell.
Založnik:
Elsevier, International Federation of Automatic Control
ISSN:
0952-1976
COBISS.SI-ID:
25396224
Licence
Licenca:
CC BY-NC 4.0, Creative Commons Priznanje avtorstva-Nekomercialno 4.0 Mednarodna
Povezava:
http://creativecommons.org/licenses/by-nc/4.0/deed.sl
Opis:
Licenca Creative Commons, ki prepoveduje komercialno uporabo, vendar uporabniki ne rabijo upravljati materialnih avtorskih pravic na izpeljanih delih z enako licenco.
Sekundarni jezik
Jezik:
Slovenski jezik
Ključne besede:
sinteza slik
,
biometrija na podlagi obraza
,
zasebni podatki
,
večspektralno razpoznavanje
,
generativne nasprotniške mreže
Projekti
Financer:
ARIS - Javna agencija za znanstvenoraziskovalno in inovacijsko dejavnost Republike Slovenije
Številka projekta:
P2-0250
Naslov:
Metrologija in biometrični sistemi
Financer:
ARIS - Javna agencija za znanstvenoraziskovalno in inovacijsko dejavnost Republike Slovenije
Številka projekta:
P2-0214
Naslov:
Računalniški vid
Financer:
ARIS - Javna agencija za znanstvenoraziskovalno in inovacijsko dejavnost Republike Slovenije
Številka projekta:
J2-50065
Naslov:
Odkrivanje globokih ponaredkov z metodami zaznave anomalij (DeepFake DAD)
Financer:
ARIS - Javna agencija za znanstvenoraziskovalno in inovacijsko dejavnost Republike Slovenije
Program financ.:
Young researchers
Podobna dela
Podobna dela v RUL:
Podobna dela v drugih slovenskih zbirkah:
Nazaj