This work discusses a CMOS-based implementation of low power speaker identification (SI) using Gaussian mixture models (GMMs). Conventionally, GMM-based SI relies on repeated access to log-add look-up table (LUT). With increasing dimensional of speaker models mathbf{and}/mathbf{or} number of speakers in the database, accesses to the LUT dominate the overall energy expense for SI. In this work, we discuss piece-wise linear approximations to GMM model that eliminate LUT accesses, thereby limiting model parameter storage in the register files alone, while incurring a minimal accuracy drop. We evaluate our scheme on TIMIT corpus where for text-independent SI and with a two-second test speech, our scheme achieves more than 90% accuracy across test-sets. We discuss the detailed architecture of the control unit, datapath, and key modules in our scheme. Compared to an equivalent design that requires LUT accesses in mathrm{off} -chip memories, our scheme limits power dissipation for SI to sim 600 mu mathbf{W} and consumes 4 imes less energy,

Low Power Speaker Identification using Look Up-free Gaussian Mixture Model in CMOS / Gianelli, A.; Iliev, N.; Nasrin, S.; Graziano, M.; Trivedi, A. R.. - ELETTRONICO. - (2019), pp. 1-3. ((Intervento presentato al convegno 22nd IEEE Symposium on Low-Power and High-Speed Chips and Systems, COOL CHIPS 2019 tenutosi a Yokohama Joho Bunka Center, Yokohama Media and Communications Center, jpn nel 2019 [10.1109/CoolChips.2019.8721354].

Low Power Speaker Identification using Look Up-free Gaussian Mixture Model in CMOS

Graziano M.;
2019

Abstract

This work discusses a CMOS-based implementation of low power speaker identification (SI) using Gaussian mixture models (GMMs). Conventionally, GMM-based SI relies on repeated access to log-add look-up table (LUT). With increasing dimensional of speaker models mathbf{and}/mathbf{or} number of speakers in the database, accesses to the LUT dominate the overall energy expense for SI. In this work, we discuss piece-wise linear approximations to GMM model that eliminate LUT accesses, thereby limiting model parameter storage in the register files alone, while incurring a minimal accuracy drop. We evaluate our scheme on TIMIT corpus where for text-independent SI and with a two-second test speech, our scheme achieves more than 90% accuracy across test-sets. We discuss the detailed architecture of the control unit, datapath, and key modules in our scheme. Compared to an equivalent design that requires LUT accesses in mathrm{off} -chip memories, our scheme limits power dissipation for SI to sim 600 mu mathbf{W} and consumes 4 imes less energy,
978-1-7281-1749-2
File in questo prodotto:
File Dimensione Formato  
Low_Power_Speaker_Identification_using_Look_Up-free_Gaussian_Mixture_Model_in_CMOS.pdf

non disponibili

Tipologia: 2a Post-print versione editoriale / Version of Record
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 811.32 kB
Formato Adobe PDF
811.32 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

Caricamento pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2919373