Sorry, you need to enable JavaScript to visit this website.

Speaker Recognition and Characterization (SPE-SPKR)

Text-dependent Speaker Verification and RSR2015 Speech Corpus


RSR2015 (Robust Speaker Recognition 2015) is the largest publicly available speech corpus for text-dependent robust speaker recognition. The current release includes 151 hours of short duration utterances spoken by 300 speakers. RSR2015 is developed by the Human Language Technology (HLT) department at Institute for Infocomm Research (I2R) in Singapore. This newsletter describes RSR2015 corpus that addresses the reviving interest of text-dependent speaker recognition.

RSR2015_v2.pdf

PDF icon RSR2015_v2.pdf (441 downloads)

Paper Details

Authors:
Submitted On:
23 February 2016 - 1:44pm
Short Link:
Type:

Document Files

RSR2015_v2.pdf

(441 downloads)

Keywords

Subscribe

[1] , "Text-dependent Speaker Verification and RSR2015 Speech Corpus", IEEE SigPort, 2014. [Online]. Available: http://sigport.org/54. Accessed: Jul. 27, 2017.
@article{54-14,
url = {http://sigport.org/54},
author = { },
publisher = {IEEE SigPort},
title = {Text-dependent Speaker Verification and RSR2015 Speech Corpus},
year = {2014} }
TY - EJOUR
T1 - Text-dependent Speaker Verification and RSR2015 Speech Corpus
AU -
PY - 2014
PB - IEEE SigPort
UR - http://sigport.org/54
ER -
. (2014). Text-dependent Speaker Verification and RSR2015 Speech Corpus. IEEE SigPort. http://sigport.org/54
, 2014. Text-dependent Speaker Verification and RSR2015 Speech Corpus. Available at: http://sigport.org/54.
. (2014). "Text-dependent Speaker Verification and RSR2015 Speech Corpus." Web.
1. . Text-dependent Speaker Verification and RSR2015 Speech Corpus [Internet]. IEEE SigPort; 2014. Available from : http://sigport.org/54

I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS


i-Vector feature representation with probabilistic linear discriminant analysis (PLDA) scoring in speaker recognition system has recently achieved effective performance even on channel mismatch conditions. In general, experiments carried out using this combined strategy employ linear discriminant analysis (LDA) after the i-Vector extraction phase to suppress irrelevant directions, such as those introduced by noise or channel distortions. However, speaker-related and -non-related variability present in the data may prevent LDA from finding the best projection matrix.

Paper Details

Authors:
Fahimeh Bahmaninezhad, John H.L. Hansen
Submitted On:
25 March 2017 - 2:33am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Poster-Slides

(53 downloads)

Paper

(98 downloads)

Keywords

Subscribe

[1] Fahimeh Bahmaninezhad, John H.L. Hansen, "I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1744. Accessed: Jul. 27, 2017.
@article{1744-17,
url = {http://sigport.org/1744},
author = {Fahimeh Bahmaninezhad; John H.L. Hansen },
publisher = {IEEE SigPort},
title = {I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS},
year = {2017} }
TY - EJOUR
T1 - I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS
AU - Fahimeh Bahmaninezhad; John H.L. Hansen
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1744
ER -
Fahimeh Bahmaninezhad, John H.L. Hansen. (2017). I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS. IEEE SigPort. http://sigport.org/1744
Fahimeh Bahmaninezhad, John H.L. Hansen, 2017. I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS. Available at: http://sigport.org/1744.
Fahimeh Bahmaninezhad, John H.L. Hansen. (2017). "I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS." Web.
1. Fahimeh Bahmaninezhad, John H.L. Hansen. I-VECTOR/PLDA SPEAKER RECOGNITION USING SUPPORT VECTORS WITH DISCRIMINANT ANALYSIS [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1744

APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK


We propose a method to improve speaker verification performance when a test utterance is very short. In some situations with short test utterances, performance of i-vector/probabilistic linear discriminant analysis systems degrades. The proposed method transforms short-utterance feature vectors to adequate vectors using a deep neural network, which compensate for short utterances.

poster.pdf

PDF icon poster.pdf (272 downloads)

Paper Details

Authors:
IL-Ho Yang, Hee-Soo Heo, Sung-Hyun Yoon, and Ha-Jin Yu
Submitted On:
8 March 2017 - 11:53pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:

Document Files

poster.pdf

(272 downloads)

Keywords

Subscribe

[1] IL-Ho Yang, Hee-Soo Heo, Sung-Hyun Yoon, and Ha-Jin Yu, "APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1715. Accessed: Jul. 27, 2017.
@article{1715-17,
url = {http://sigport.org/1715},
author = {IL-Ho Yang; Hee-Soo Heo; Sung-Hyun Yoon; and Ha-Jin Yu },
publisher = {IEEE SigPort},
title = {APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK},
year = {2017} }
TY - EJOUR
T1 - APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK
AU - IL-Ho Yang; Hee-Soo Heo; Sung-Hyun Yoon; and Ha-Jin Yu
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1715
ER -
IL-Ho Yang, Hee-Soo Heo, Sung-Hyun Yoon, and Ha-Jin Yu. (2017). APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK. IEEE SigPort. http://sigport.org/1715
IL-Ho Yang, Hee-Soo Heo, Sung-Hyun Yoon, and Ha-Jin Yu, 2017. APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK. Available at: http://sigport.org/1715.
IL-Ho Yang, Hee-Soo Heo, Sung-Hyun Yoon, and Ha-Jin Yu. (2017). "APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK." Web.
1. IL-Ho Yang, Hee-Soo Heo, Sung-Hyun Yoon, and Ha-Jin Yu. APPLYING COMPENSATION TECHNIQUES ON I-VECTORS EXTRACTED FROM SHORT-TEST UTTERANCES FOR SPEAKER VERIFICATION USING DEEP NEURAL NETWORK [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1715

DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS

Paper Details

Authors:
Rosanna Milner, Thomas Hain
Submitted On:
7 March 2017 - 8:40am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

talk-dia-icassp17-milner.pdf

(46 downloads)

talk-dia-icassp17-milner.pdf

(48 downloads)

Keywords

Additional Categories

Subscribe

[1] Rosanna Milner, Thomas Hain, "DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1677. Accessed: Jul. 27, 2017.
@article{1677-17,
url = {http://sigport.org/1677},
author = {Rosanna Milner; Thomas Hain },
publisher = {IEEE SigPort},
title = {DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS},
year = {2017} }
TY - EJOUR
T1 - DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS
AU - Rosanna Milner; Thomas Hain
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1677
ER -
Rosanna Milner, Thomas Hain. (2017). DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS. IEEE SigPort. http://sigport.org/1677
Rosanna Milner, Thomas Hain, 2017. DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS. Available at: http://sigport.org/1677.
Rosanna Milner, Thomas Hain. (2017). "DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS." Web.
1. Rosanna Milner, Thomas Hain. DNN APPROACH TO SPEAKER DIARISATION USING SPEAKER CHANNELS [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1677

SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS


A novel speaker segmentation approach based on deep neural network is proposed and investigated. This approach uses deep speaker vectors (d-vectors) to represent speaker characteristics and to find speaker change points. The d-vector is a kind of frame-level speaker recognition feature, whose discriminative training process corresponds to the goal of discriminating a speaker change point from a single speaker speech segment in a short time window.

Paper Details

Authors:
Submitted On:
28 February 2017 - 4:11am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS

(54 downloads)

Keywords

Subscribe

[1] , "SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1490. Accessed: Jul. 27, 2017.
@article{1490-17,
url = {http://sigport.org/1490},
author = { },
publisher = {IEEE SigPort},
title = {SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS},
year = {2017} }
TY - EJOUR
T1 - SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS
AU -
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1490
ER -
. (2017). SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS. IEEE SigPort. http://sigport.org/1490
, 2017. SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS. Available at: http://sigport.org/1490.
. (2017). "SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS." Web.
1. . SPEAKER SEGMENTATION USING DEEP SPEAKER VECTORS FOR FAST SPEAKER CHANGE SCENARIOS [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1490

EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION


The universal speech attributes for speaker verification (SV)
are addressed in this paper. The aim of this work is to
exploit fundamental characteristics across different speakers
within the deep neural network (DNN)/i-vector framework.
The manner and place of articulation form the fundamental
speech attribute unit inventory, and new attribute units for
acoustic modelling are generated by a two-step automatic
clustering method in this paper. The DNN based on
universal attribute units is used to generate posterior

Paper Details

Authors:
Sheng Zhang, Wu Guo, Guoping Hu
Submitted On:
27 February 2017 - 8:54pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP2017_shengzhang_v2.pdf

(62 downloads)

Keywords

Subscribe

[1] Sheng Zhang, Wu Guo, Guoping Hu, "EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1462. Accessed: Jul. 27, 2017.
@article{1462-17,
url = {http://sigport.org/1462},
author = {Sheng Zhang; Wu Guo; Guoping Hu },
publisher = {IEEE SigPort},
title = {EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION},
year = {2017} }
TY - EJOUR
T1 - EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION
AU - Sheng Zhang; Wu Guo; Guoping Hu
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1462
ER -
Sheng Zhang, Wu Guo, Guoping Hu. (2017). EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION. IEEE SigPort. http://sigport.org/1462
Sheng Zhang, Wu Guo, Guoping Hu, 2017. EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION. Available at: http://sigport.org/1462.
Sheng Zhang, Wu Guo, Guoping Hu. (2017). "EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION." Web.
1. Sheng Zhang, Wu Guo, Guoping Hu. EXPLORING UNIVERSAL SPEECH ATTRIBUTES FOR SPEAKER VERIFICATION [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1462

Senone I-Vectors for Robust Speaker Verification

Paper Details

Authors:
Submitted On:
15 October 2016 - 7:50am
Short Link:
Type:
Event:
Presenter's Name:

Document Files

ISCSLP16_SenoneIvector.pdf

(86 downloads)

Keywords

Subscribe

[1] , "Senone I-Vectors for Robust Speaker Verification", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1219. Accessed: Jul. 27, 2017.
@article{1219-16,
url = {http://sigport.org/1219},
author = { },
publisher = {IEEE SigPort},
title = {Senone I-Vectors for Robust Speaker Verification},
year = {2016} }
TY - EJOUR
T1 - Senone I-Vectors for Robust Speaker Verification
AU -
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1219
ER -
. (2016). Senone I-Vectors for Robust Speaker Verification. IEEE SigPort. http://sigport.org/1219
, 2016. Senone I-Vectors for Robust Speaker Verification. Available at: http://sigport.org/1219.
. (2016). "Senone I-Vectors for Robust Speaker Verification." Web.
1. . Senone I-Vectors for Robust Speaker Verification [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1219

Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences


The widely adopted i-vector performances well in text-independent speaker verification with long speech duration. How to integrate the state-of-the-art i-vector framework into the text-prompted speaker verification is addressed in this paper. To take advantage of the lexical information and enhance the performance for speaker verification with random digit sequences, this paper proposes to extract a set of digit-dependent local i-vectors from the utterance instead of extracting a single i-vector. The digit-dependent local i-vector is considered

Paper Details

Authors:
Peixin Chen, Wu Guo, Guoping Hu
Submitted On:
14 October 2016 - 10:24pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ISCSLP2016_PeixinChen.pdf

(89 downloads)

Keywords

Subscribe

[1] Peixin Chen, Wu Guo, Guoping Hu, "Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1213. Accessed: Jul. 27, 2017.
@article{1213-16,
url = {http://sigport.org/1213},
author = {Peixin Chen; Wu Guo; Guoping Hu },
publisher = {IEEE SigPort},
title = {Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences},
year = {2016} }
TY - EJOUR
T1 - Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences
AU - Peixin Chen; Wu Guo; Guoping Hu
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1213
ER -
Peixin Chen, Wu Guo, Guoping Hu. (2016). Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences. IEEE SigPort. http://sigport.org/1213
Peixin Chen, Wu Guo, Guoping Hu, 2016. Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences. Available at: http://sigport.org/1213.
Peixin Chen, Wu Guo, Guoping Hu. (2016). "Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences." Web.
1. Peixin Chen, Wu Guo, Guoping Hu. Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with Random Digit Sequences [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1213

Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with


The widely adopted i-vector performances well in textindependent speaker verification with long speech duration.

Paper Details

Authors:
Peixin Chen, Wu Guo, Guoping Hu
Submitted On:
14 October 2016 - 10:24pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ISCSLP2016_PeixinChen.pdf

(96 downloads)

Keywords

Subscribe

[1] Peixin Chen, Wu Guo, Guoping Hu, "Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1206. Accessed: Jul. 27, 2017.
@article{1206-16,
url = {http://sigport.org/1206},
author = {Peixin Chen; Wu Guo; Guoping Hu },
publisher = {IEEE SigPort},
title = {Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with},
year = {2016} }
TY - EJOUR
T1 - Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with
AU - Peixin Chen; Wu Guo; Guoping Hu
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1206
ER -
Peixin Chen, Wu Guo, Guoping Hu. (2016). Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with. IEEE SigPort. http://sigport.org/1206
Peixin Chen, Wu Guo, Guoping Hu, 2016. Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with. Available at: http://sigport.org/1206.
Peixin Chen, Wu Guo, Guoping Hu. (2016). "Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with." Web.
1. Peixin Chen, Wu Guo, Guoping Hu. Digit-dependent Local I-Vector for Text-Prompted Speaker Verification with [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1206

A study of variational method for text-independent speaker recognition


An i-vector has become the state-of-the-art algorithm for text-independent recognition. Most of related works take the extraction of the i-vector as a black-box by using some open software (e.g. Kaldi, Alize) and focus on the vector-based back-end algorithms, such as length normalization, WCCN, or PLDA. In this paper, we study the variational method and present a concise derivation for the i-vector. Based on our proposed methods, three criteria for derivation are compared. There are maximum likelihood (ML), maximum a posteriori (MAP) and maximum

Paper Details

Authors:
Liang He, Yao Tian, Yi Liu, Fang Dong, WeiQiang Zhang, Jia Liu
Submitted On:
13 October 2016 - 11:19pm
Short Link:
Type:
Event:

Document Files

The poster in ISCSLP2016

(83 downloads)

Keywords

Subscribe

[1] Liang He, Yao Tian, Yi Liu, Fang Dong, WeiQiang Zhang, Jia Liu, "A study of variational method for text-independent speaker recognition", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1179. Accessed: Jul. 27, 2017.
@article{1179-16,
url = {http://sigport.org/1179},
author = {Liang He; Yao Tian; Yi Liu; Fang Dong; WeiQiang Zhang; Jia Liu },
publisher = {IEEE SigPort},
title = {A study of variational method for text-independent speaker recognition},
year = {2016} }
TY - EJOUR
T1 - A study of variational method for text-independent speaker recognition
AU - Liang He; Yao Tian; Yi Liu; Fang Dong; WeiQiang Zhang; Jia Liu
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1179
ER -
Liang He, Yao Tian, Yi Liu, Fang Dong, WeiQiang Zhang, Jia Liu. (2016). A study of variational method for text-independent speaker recognition. IEEE SigPort. http://sigport.org/1179
Liang He, Yao Tian, Yi Liu, Fang Dong, WeiQiang Zhang, Jia Liu, 2016. A study of variational method for text-independent speaker recognition. Available at: http://sigport.org/1179.
Liang He, Yao Tian, Yi Liu, Fang Dong, WeiQiang Zhang, Jia Liu. (2016). "A study of variational method for text-independent speaker recognition." Web.
1. Liang He, Yao Tian, Yi Liu, Fang Dong, WeiQiang Zhang, Jia Liu. A study of variational method for text-independent speaker recognition [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1179

Pages