Learnable MFCCs for Speaker Verification

被引：5

作者：

Liu, Xuechen ^{[1
,2
]}

Sahidullah, Md ^{[2
]}

Kinnunen, Tomi ^{[1
]}

机构：

[1] Univ Eastern Finland, Sch Comp, Joensuu, Finland

[2] Univ Lorraine, CNRS, INRIA, LORIA, F-54000 Nancy, France

来源：

2021 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS (ISCAS) | 2021年

基金：

芬兰科学院;

关键词：

Speaker verification; feature extraction; mel-frequency cesptral coefficients (MFCCs); RECOGNITION; FEATURES;

D O I：

10.1109/ISCAS51556.2021.9401593

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

We propose a learnable mel-frequency cepstral coefficients (MFCCs) front-end architecture for deep neural network (DNN) based automatic speaker verification. Our architecture retains the simplicity and interpretability of MFCC-based features while allowing the model to be adapted to data flexibly. In practice, we formulate data-driven version of four linear transforms in a standard MFCC extractor - windowing, discrete Fourier transform (DFT), mel filterbank and discrete cosine transform (DCT). Results reported reach up to 6.7% (VoxCeleb1) and 9.7% (SITW) relative improvement in term of equal error rate (EER) from static MFCCs, without additional tuning effort.

引用

页数：5

共 50 条

[21] SPEAKER VERIFICATION - A TUTORIAL
NAIK, JM
IEEE COMMUNICATIONS MAGAZINE, 1990, 28 (01) : 42 - 48
[22] Disentangling speaker and channel effects in speaker verification
Kenny, P
Dumouchel, P
2004 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, VOL I, PROCEEDINGS: SPEECH PROCESSING, 2004, : 37 - 40
[23] Deep Speaker Embeddings for Speaker Verification of Children
Abed, Mohammed Hamzah
Sztaho, David
TEXT, SPEECH, AND DIALOGUE, TSD 2024, PT II, 2024, 15049 : 58 - 69
[24] DISENTANGLED SPEAKER EMBEDDING FOR ROBUST SPEAKER VERIFICATION
Yi, Lu
Mak, Man-Wai
2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2022, : 7662 - 7666
[25] Speaker verification without background speaker models
Hsu, CN
Yu, HC
Yang, BH
2003 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, VOL II, PROCEEDINGS: SPEECH II; INDUSTRY TECHNOLOGY TRACKS; DESIGN & IMPLEMENTATION OF SIGNAL PROCESSING SYSTEMS; NEURAL NETWORKS FOR SIGNAL PROCESSING, 2003, : 233 - 236
[26] SPEAKER VERIFICATION BY INEXPERIENCED AND EXPERIENCED LISTENERS VS. SPEAKER VERIFICATION SYSTEM
Kahn, Juliette
Audibert, Nicolas
Rossato, Solange
Bonastre, Jean-Francois
2011 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2011, : 5912 - 5915
[27] Deep neural network framework and transformed MFCCs for speaker's age and gender classification
Qawaqneh, Zakariya
Abu Mallouh, Arafat
Barkana, Buket D.
KNOWLEDGE-BASED SYSTEMS, 2017, 115 : 5 - 14
[28] Low-SNR, Speaker-Dependent Speech Enhancement using GMMs and MFCCs
Boucheron, Laura E.
De Leon, Phillip L.
13TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2012 (INTERSPEECH 2012), VOLS 1-3, 2012, : 574 - 577
[29] Introducing phonetic information to speaker embedding for speaker verification
Liu, Yi
He, Liang
Liu, Jia
Johnson, Michael T.
EURASIP JOURNAL ON AUDIO SPEECH AND MUSIC PROCESSING, 2019, 2019 (01)
[30] TOWARDS ROBUST SPEAKER VERIFICATION WITH TARGET SPEAKER ENHANCEMENT
Zhang, Chunlei
Yu, Meng
Weng, Chao
Yu, Dong
2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021), 2021, : 6693 - 6697

← 1 2 3 4 5 →