In the last decade, demographic profiling from facial imagery has grown in its importance in the computer vision field. For demographic profiling, we usually mean gender, ethnicity, and age identification from face images. In this paper, we propose an efficient and effective profiling framework and we assess the quality of the proposed approach comparing the results obtained by our system with those achieved by other recently published methods on large datasets of facial images with different age, gender, and ethnicity. These results show how a carefully engineered pipeline of efficient image analysis and pattern recognition techniques leads to state-of-the-art results at 20FPS using a single thread on a 1.6GHZ i5-2467M processor.
Real-Time Demographic Profiling from Face Imagery with Fisher Vectors / Lorenzo Seidenari, Alessandro Rozza, Alberto Del Bimbo. - In: MACHINE VISION AND APPLICATIONS. - ISSN 0932-8092. - ELETTRONICO. - (2019), pp. 1-16. [10.1007/s00138-018-0991-2]
Real-Time Demographic Profiling from Face Imagery with Fisher Vectors
Lorenzo Seidenari
;Alberto Del Bimbo
2019
Abstract
In the last decade, demographic profiling from facial imagery has grown in its importance in the computer vision field. For demographic profiling, we usually mean gender, ethnicity, and age identification from face images. In this paper, we propose an efficient and effective profiling framework and we assess the quality of the proposed approach comparing the results obtained by our system with those achieved by other recently published methods on large datasets of facial images with different age, gender, and ethnicity. These results show how a carefully engineered pipeline of efficient image analysis and pattern recognition techniques leads to state-of-the-art results at 20FPS using a single thread on a 1.6GHZ i5-2467M processor.File | Dimensione | Formato | |
---|---|---|---|
realtime-age-estimation-mva.pdf
Accesso chiuso
Tipologia:
Versione finale referata (Postprint, Accepted manuscript)
Licenza:
Tutti i diritti riservati
Dimensione
1.21 MB
Formato
Adobe PDF
|
1.21 MB | Adobe PDF | Richiedi una copia |
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.