TU Darmstadt / ULB / TUbiblio

How Colorful Should Faces Be? Harmonizing Color and Model Quantization for Resource-restricted Face Recognition

Kolf, Jan Niklas ; Elliesen, Jurek ; Boutros, Fadi ; Damer, Naser (2023)
How Colorful Should Faces Be? Harmonizing Color and Model Quantization for Resource-restricted Face Recognition.
International Joint Conference on Biometrics 2023. Ljubljana, Slovenia (25.-28.9.2023)
doi: 10.1109/IJCB57857.2023.10449031
Konferenzveröffentlichung, Bibliographie

Kurzbeschreibung (Abstract)

State-of-the-art face recognition (FR) systems are based on overparameterized deep neural networks (DNN) which commonly use face images with 256 3 colors. The use of DNN and the storage of face images as references for comparison are limited in resource-restricted domains, which are hemmed in storage and computational capacity. A possible solution is to store the image only as a feature, which renders the human evaluation of the image impossible and forces the use of a single DNN (vendor) across systems. In this paper, we present a novel study on the possibility and effect of image color quantization on FR performance and storage efficiency. We leverage our conclusions to propose harmonizing the color quantization with the low-bit quantization of FR models. This combination significantly reduces the bits required to represent both the image and the FR model. In an extensive experiment on diverse sets of DNN architectures and color quantization steps, we validate on multiple benchmarks that the proposed methodology can successfully reduce the number of bits required for image pixels and DNN data while maintaining nearly equal recognition rates. The code and pre-trained models are available at https://github.com/jankolf/ColorQuantization.

Typ des Eintrags: Konferenzveröffentlichung
Erschienen: 2023
Autor(en): Kolf, Jan Niklas ; Elliesen, Jurek ; Boutros, Fadi ; Damer, Naser
Art des Eintrags: Bibliographie
Titel: How Colorful Should Faces Be? Harmonizing Color and Model Quantization for Resource-restricted Face Recognition
Sprache: Englisch
Publikationsjahr: 29 September 2023
Verlag: IEEE
Buchtitel: 2023 IEEE International Joint Conference on Biometrics (IJCB)
Veranstaltungstitel: International Joint Conference on Biometrics 2023
Veranstaltungsort: Ljubljana, Slovenia
Veranstaltungsdatum: 25.-28.9.2023
DOI: 10.1109/IJCB57857.2023.10449031
Kurzbeschreibung (Abstract):

State-of-the-art face recognition (FR) systems are based on overparameterized deep neural networks (DNN) which commonly use face images with 256 3 colors. The use of DNN and the storage of face images as references for comparison are limited in resource-restricted domains, which are hemmed in storage and computational capacity. A possible solution is to store the image only as a feature, which renders the human evaluation of the image impossible and forces the use of a single DNN (vendor) across systems. In this paper, we present a novel study on the possibility and effect of image color quantization on FR performance and storage efficiency. We leverage our conclusions to propose harmonizing the color quantization with the low-bit quantization of FR models. This combination significantly reduces the bits required to represent both the image and the FR model. In an extensive experiment on diverse sets of DNN architectures and color quantization steps, we validate on multiple benchmarks that the proposed methodology can successfully reduce the number of bits required for image pixels and DNN data while maintaining nearly equal recognition rates. The code and pre-trained models are available at https://github.com/jankolf/ColorQuantization.

Freie Schlagworte: Biometrics, Machine learning, Efficiency, Face recognition, Deep learning
Fachbereich(e)/-gebiet(e): 20 Fachbereich Informatik
20 Fachbereich Informatik > Graphisch-Interaktive Systeme
Hinterlegungsdatum: 12 Apr 2024 10:27
Letzte Änderung: 12 Apr 2024 10:30
PPN:
Export:
Suche nach Titel in: TUfind oder in Google
Frage zum Eintrag Frage zum Eintrag

Optionen (nur für Redakteure)
Redaktionelle Details anzeigen Redaktionelle Details anzeigen