VoxCeleb: A Large-Scale Speaker Identification Dataset
Open accessThe DOI matches its Crossref record; 10 of the 10 references it cites were found in a registry or an index; none of the 16 sources its claims cite is checked for support yet.
Abstract
Most existing datasets for speaker identification contain samples obtained under quite constrained conditions, and are usually hand-annotated, hence limited in size. The goal of this paper is to generate a large scale text-independent speaker identification dataset collected 'in the wild'. We make two contributions. First, we propose a fully automated pipeline based on computer vision techniques to create the dataset from open-source media. Our pipeline involves obtaining videos from YouTube; performing active speaker verification using a two-stream synchronization Convolutional Neural Network (CNN), and confirming the identity of the speaker using CNN based facial recognition. We use this pipeline to curate VoxCeleb which contains hundreds of thousands of 'real world' utterances for over 1,000 celebrities. Our second contribution is to apply and compare various state of the art speaker identification techniques on our dataset to establish baseline performance. We show that a CNN based architecture obtains the best performance for both identification and verification.
Overview
CoChat Summary
TL;DR
The paper introduces VoxCeleb, a fully automated, large-scale speaker identification dataset built from YouTube videos “in the wild,” with 1,251 celebrities and over 100,000 utterances. It also benchmarks speaker recognition methods and reports that a CNN performs best on both identification and verification.
Key findings
- VoxCeleb contains over 100,000 utterances for 1,251 celebrities; the dataset is gender balanced, with 55% male speakers.
- The automated collection pipeline uses active speaker verification and face verification and was tuned to be highly precise: precision = 1.000 for both stages, with recall = 0.613 for active speaker verification and 0.726 for face verification.
- For speaker identification, the CNN achieves 80.5% top-1 and 92.1% top-5 accuracy, compared with 49.0% / 56.6% for I-vectors + SVM and 60.8% / 75.6% for I-vectors + PLDA + SVM.
- For speaker verification, the best reported result is CNN + Embedding with C min det = 0.71 and EER = 7.8%, better than I-vectors + PLDA (0.73, 8.8%) and GMM-UBM (0.80, 15.0%).
Methods
- Built VoxCeleb from YouTube videos using a fully automated multi-stage pipeline: candidate POI list, video download, face tracking, active speaker verification via SyncNet, and face verification via VGG Face CNN.
- Dataset scale/statistics reported: 1,251 POIs, 690 male POIs, 36 / 18 / 8 videos per POI (max/avg/min), 250 / 123 / 45 utterances per POI (max/avg/min), and 145.0 / 8.2 / 4.0 s utterance length (max/avg/min).
- Trained and evaluated a CNN on spectrograms with minimal preprocessing; identification used a 1,251-way classification setup, while verification compared I-vectors/PLDA, GMM-UBM, and CNN embedding approaches.
Limitations
- The paper states that the pipeline is designed to be conservative, and that this means “a lot of the downloaded videos” are discarded to reduce label errors.
- The paper notes that the collection starts from a list of 2,622 identities from VGG Face and uses YouTube search with “interview” appended, which is an evident source of selection bias toward publicly available, interview-like media.
- The dataset is described as coming from open-source media / YouTube, so it is limited to speakers and contexts available there; this is evident from the text, though not stated as a formal limitation.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 1 Not valid · 12 Not checked · 37 Own results and interpretations
In the paper
“| Face verification | [30] | 1.000 | 0.726 |”
What the sources say
- Not checked yetcites [30] [30] O. M. Parkhi, A. Vedaldi, and A. Zisserman, 'Deep face recognition,' in Proceedings of the British Machine Vision Conference , 2015. -
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Unlike the SITW dataset, both audio and video for each speaker is released.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“| Active speaker verification | [31] | 1.000 | 0.613 |”
What the sources say
- Not checked yetcites [31] [31] P. Chakravarty and T. Tuytelaars, 'Cross-modal supervision for learning active speaker detection in video,' arXiv preprint arXiv:1603.08907 , 2016. -
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“reduces the number of parameters from 319M in VGG-M to 67M in our network”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“| # of videos per POI | 36 / 18 / 8 |”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
15 more claims
Create a free account to see all 50 claims and to verify the 16 references still open.
- Probably not a substantial claim. 9
- It may state two separate claims as one. 23
- The quote only partly states this claim. 5
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Meanwhile, the emergence of large-scale VSA datasets [28, 29, 30, 31, 32] promoted the further development of deep learning based VSA research.”
From Deep learning for visual speech analysis: A survey (2024) · cites this paper as [31]
In the citing paper
“Although it is mainly built for speaker identification, it also can be used for VSG.”
From Deep learning for visual speech analysis: A survey (2024) · cites this paper as [31]
In the citing paper
“It contains over 100k utterances from 1,251 celebrities.”
From Deep learning for visual speech analysis: A survey (2024) · cites this paper as [31]
In the citing paper
“VoxCeleb1 [31] is a large-scale text-independent audio-visual dataset collected from open-source YouTube media.”
From Deep learning for visual speech analysis: A survey (2024) · cites this paper as [31]
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818 · cited 1×
“The network is trained using batch normalisation [40] and all hyper-parameters (e.”
MethodBatch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S. (2024) · arXiv (Cornell University) · doi:10.57702/o9raffed · Cited by 14,214
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1512.03385 · Cited by 4,196
One millisecond face alignment with an ensemble of regression trees
Kazemi, V., & Sullivan, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.241 · Cited by 2,831 · cited 1×
“Facial landmark positions are detected for each face detection using the regression tree based method of [33].”
MethodCNN architectures for large-scale audio classification
Hershey, S., Chaudhuri, S., Ellis, D. P. W., Gemmeke, J. F., Jansen, A., Moore, R. C., Plakal, M., Platt, D., Saurous, R. A., Seybold, B., et al. (2017) · 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2017.7952132 · Cited by 2,023
MatConvNet - Convolutional Neural Networks for MATLAB
Vedaldi, A., & Lenc, K. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1412.4564 · Cited by 578 · cited 1×
“Our implementation is based on the deep learning toolbox MatConvNet [39] and trained on a NVIDIA TITAN X GPU.”
MethodCNN Architectures for Large-Scale Audio ClassificationPreprint
Hershey, S., Chaudhuri, S., Ellis, D. P. W., Gemmeke, J. F., Jansen, A., Moore, R. C., Plakal, M., Platt, D., Saurous, R. A., Seybold, B., et al. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1609.09430 · Cited by 217 · cited 1×
“This has led to a very recent shift from hand-crafted features to the domain of deep CNNs which can be applied to higher dimensional inputs [13, 14] and for speaker iden-tification [15].”
BackgroundCross-Modal Supervision for Learning Active Speaker Detection in Video
Chakravarty, P., & Tuytelaars, T. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-46454-1_18 · Cited by 70 · cited 2×
“Precision-recall curves for both tasks on their respective benchmark datasets [30, 31] are shown in Figure 1 right, and the values at the operating point are given in Table 3.”
BackgroundCross-modal Supervision for Learning Active Speaker Detection in VideoPreprint
Chakravarty, P., & Tuytelaars, T. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1603.08907 · Cited by 1 · cited 2×
“Precision-recall curves for both tasks on their respective benchmark datasets [30, 31] are shown in Figure 1 right, and the values at the operating point are given in Table 3.”
Background
Research works printed but not linked: 28
- [2][2] A. Krizhevsky, I. Sutskever, and G. E. Hinton, 'ImageNet classification with deep convolutional neural networks,' in Advances in Neural Information Processing Systems , pp. 1106-1114, 2012. -
- [5][5] M. McLaren, L. Ferrer, D. Castan, and A. Lawson, 'The speakers in the wild (SITW) speaker recognition database,' INTERSPEECH , vol. 2016, 2016. -
- [6][6] O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, S. Huang, A. Karpathy, A. Khosla, M. Bernstein, A. Berg, and F. Li, 'Imagenet large scale visual recognition challenge,' International Journal of Computer Vision , 2015. -
- [7][7] D. A. Reynolds, T. F. Quatieri, and R. B. Dunn, 'Speaker verification using adapted gaussian mixture models,' Digital signal processing , vol. 10, no. 1-3, pp. 19-41, 2000. -
- [8][8] D. A. Reynolds and R. C. Rose, 'Robust text-independent speaker identification using gaussian mixture speaker models,' IEEE transactions on speech and audio processing , vol. 3, no. 1, pp. 7283, 1995. -
- [10][10] N. Dehak, P. J. Kenny, R. Dehak, P. Dumouchel, and P. Ouellet, 'Front-end factor analysis for speaker verification,' IEEE Transactions on Audio, Speech, and Language Processing , vol. 19, no. 4, pp. 788-798, 2011. -
- [11][11] U. H. Yapanel, X. Zhang, and J. H. Hansen, 'High performance digit recognition in real car environments.,' in INTERSPEECH , 2002. -
- [12][12] J. H. Hansen, R. Sarikaya, U. H. Yapanel, and B. L. Pellom, 'Robust speech recognition in noise: an evaluation using the spine corpus.,' in INTERSPEECH , pp. 905-908, 2001. -
- [13][13] T. N. Sainath, R. J. Weiss, A. W. Senior, K. W. Wilson, and O. Vinyals, 'Learning the speech front-end with raw waveform CLDNNs,' in INTERSPEECH , pp. 1-5, 2015. -
- [15][15] Y. Lukic, C. Vogt, O. D¨ urr, and T. Stadelmann, 'Speaker identification and clustering using convolutional neural networks,' in IEEE 26th International Workshop on Machine Learning for Signal Processing (MLSP) , pp. 1-6, IEEE, 2016. -
- [16][16] D. van der Vloed, J. Bouten, and D. A. van Leeuwen, 'NFIFRITS: a forensic speaker recognition database and some first experiments,' in The Speaker and Language Recognition Workshop , 2014. -
- [17][17] J. Hennebert, H. Melin, D. Petrovska, and D. Genoud, 'POLYCOST: a telephone-speech database for speaker recognition,' Speech communication , vol. 31, no. 2, pp. 265-270, 2000. -
- [18][18] J. B. Millar, J. P. Vonwiller, J. M. Harrington, and P. J. Dermody, 'The Australian national database of spoken language,' in Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing , vol. 1, pp. I-97, IEEE, 1994. -
- [19][19] J. S. Garofolo, L. F. Lamel, W. M. Fisher, J. G. Fiscus, and D. S. Pallett, 'DARPA TIMIT acoustic-phonetic continous speech corpus CD-ROM. NIST speech disc 1-1.1,' NASA STI/Recon technical report , vol. 93, 1993. -
- [21][21] R. Woo, A. Park, and T. J. Hazen, 'The MIT Mobile Device Speaker Verification Corpus: Data collection and preliminary experiments,' The Speaker and Language Recognition Workshop , 2006. -
- [23][23] A. Janin, D. Baron, J. Edwards, D. Ellis, D. Gelbart, N. Morgan, B. Peskin, T. Pfau, E. Shriberg, A. Stolcke, et al. , 'The ICSI meeting corpus,' in Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing , vol. 1, IEEE, 2003. -
- [24][24] I. McCowan, J. Carletta, W. Kraaij, S. Ashby, S. Bourban, M. Flynn, M. Guillemot, T. Hain, J. Kadlec, V. Karaiskos, et al. , 'The AMI meeting corpus,' in International Conference on Methods and Techniques in Behavioral Research , vol. 88, 2005. -
- [25][25] P. Bell, M. J. Gales, T. Hain, J. Kilgour, P. Lanchantin, X. Liu, A. McParland, S. Renals, O. Saz, M. Wester, et al. , 'The MGB challenge: Evaluating multi-genre broadcast media recognition,' in IEEE Workshop on Automatic Speech Recognition and Understanding , pp. 687-693, IEEE, 2015. -
- [27][27] J. J. Godfrey, E. C. Holliman, and J. McDaniel, 'Switchboard: Telephone speech corpus for research and development,' in Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing , vol. 1, pp. 517-520, IEEE, 1992. -
- [28][28] W. M. Fisher, G. R. Doddington, and K. M. Goudie-Marshall, 'The DARPA speech recognition research database: specifications and status,' in Proc. DARPA Workshop on speech recognition , pp. 93-99, 1986. -
- [30][30] O. M. Parkhi, A. Vedaldi, and A. Zisserman, 'Deep face recognition,' in Proceedings of the British Machine Vision Conference , 2015. -
- [32][32] D. E. King, 'Dlib-ml: A machine learning toolkit,' The Journal of Machine Learning Research , vol. 10, pp. 1755-1758, 2009. -
- [34][34] J. S. Chung and A. Zisserman, 'Lip reading in the wild,' in Proceedings of the Asian Conference on Computer Vision , 2016. -
- [35][35] M. Everingham, J. Sivic, and A. Zisserman, 'Taking the bite out of automatic naming of characters in TV video,' Image and Vision Computing , vol. 27, no. 5, 2009. -
- [36][36] J. S. Chung and A. Zisserman, 'Out of time: automated lip sync in the wild,' in Workshop on Multi-view Lip-reading, ACCV , 2016. -
- [37][37] K. Chatfield, K. Simonyan, A. Vedaldi, and A. Zisserman, 'Return of the devil in the details: Delving deep into convolutional nets,' in Proceedings of the British Machine Vision Conference , 2014. -
- [38][38] S. Chopra, R. Hadsell, and Y. LeCun, 'Learning a similarity metric discriminatively, with application to face verification,' in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , vol. 1, pp. 539-546, IEEE, 2005. -
- [41][41] S. Ioffe, 'Probabilistic linear discriminant analysis,' in Proceedings of the European Conference on Computer Vision , pp. 531542, Springer, 2006.
Cited by
Items 701–800 of 1,008
Deepfake audio detection and justification with explainable artificial intelligence
Govindu, A., Kale, P., Hullur, A., Gurav, A., & Godse, P. (2026) · International Journal of Data Science and Analytics · doi:10.1007/s41060-026-01186-7 · Cited by 16
“The FakeAVCeleb dataset contains manipulated audio clips of celebrities created using voice conversion techniques [64].”
Semi-supervised audio-driven TV-news speaker diarization using deep neural embeddings
Tsipas, N., Vrysis, L., Konstantoudakis, K., & Dimoulas, C. A. (2020) · The Journal of the Acoustical Society of America · doi:10.1121/10.0002924 · Cited by 12
Review of different robust x-vector extractors for speaker verification
Rouvier, M., Dufour, R., & Bousquet, P.-M. (2020) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco47968.2020.9287426 · Cited by 12
“The trained extractors are evaluated on Speakers in the Wild (SITW) core-core task [21] and Voxceleb1-E Cleaned [22] dataset with 16 KHz sampling rate.”
MethodSupervised Attention for Speaker Recognition
Kye, S. M., Chung, J. S., & Kim, H. (2021) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt48900.2021.9383579 · Cited by 12
“To show the effectiveness of our models, we train the models with various loss function on the VoxCeleb1 dataset [11].”
BackgroundMethodInfluentialSelf-citationShi, Y., Huang, Q., & Hain, T. (2021) · Neural Networks · doi:10.1016/j.neunet.2021.05.024 · Cited by 13
Deep Learning End to End Speech Synthesis: A Review
Nazir, O., & Malik, A. (2021) · International Conference on Secure Cyber Computing and Communications (ICSCCC) · doi:10.1109/icsccc51823.2021.9478125 · Cited by 12
Emotional Prosody Control for Speech Generation
Sivaprasad, S., Kosgi, S., & Gandhi, V. (2021) · Interspeech (USB) · doi:10.21437/interspeech.2021-307 · Cited by 18
“More recent approaches decouple speaker modeling from speech synthesis by independently training a speaker-discriminative embedding network [19].”
MethodKanda, N., Xiao, X., Wu, J., Zhou, T., Gaur, Y., Wang, X., Meng, Z., Chen, Z., & Yoshioka, T. (2021) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru51503.2021.9687974 · Cited by 25
Class token and knowledge distillation for multi-head self-attention speaker verification systems
Mingote, V., Miguel, A., Ortega, A., & Lleida, E. (2022) · Digital Signal Processing · doi:10.1016/j.dsp.2022.103859 · Cited by 17
Ciaparrone, G., Chiariglione, L., & Tagliaferri, R. (2022) · Neural Computing and Applications · doi:10.1007/s00521-021-06875-x · Cited by 14
“It was published in two versions: VoxCeleb1 [65] and VoxCeleb2 [7].”
MethodContrastive-mixup Learning for Improved Speaker Verification
Zhang, X., Jin, M., Cheng, R., Li, R., Han, E., & Stolcke, A. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746411 · Cited by 20
“We use the VoxCeleb1 [26] test set to evaluate model performance.”
MethodAnalyzing Language-Independent Speaker Anonymization Framework under Unseen Conditions
Miao, X., Wang, X., Cooper, E., Yamagishi, J., & Tomashenko, N. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-11065 · Cited by 19
AnimeCeleb: Large-Scale Animation CelebHeads Dataset for Head Reenactment
Kim, K., Park, S., Lee, J., Chung, S., Lee, J., & Choo, J. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20074-8_24 · Cited by 19
“3, we use both cartoon texture shader style AnimeCeleb and the VoxCeleb[14] as a training dataset.”
BackgroundMethodInfluentialEnrollment-Stage Backdoor Attacks on Speaker Recognition Systems via Adversarial Ultrasound
Li, X., Ze, J., Yan, C., Cheng, Y., Ji, X., & Xu, W. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3328253 · Cited by 20
“We validate TUNER on seven representative SRS models, including ECAPA-TDNN [15], Pyannote [16], U-Level [17], WavLM-Xvec [18], SpeakerNet [19], D-vector [20], SincXvec [21] along with two typical speech datasets (VoxCeleb1 [22] and LibriSpeech [23]).”
MethodInfluentialSpeech refinement using Bi-LSTM and improved spectral clustering in speaker diarization
Gupta, A., & Purwar, A. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-17017-x · Cited by 12
Enhancement and Analysis of Conversational Speech: JSALT 2017
Ryanta, N., Bergelson, E., Church, K., Cristià, A., Du, J., Ganapathy, S., Khudanpur, S. P., Kowalski, D., Krishnamoorthy, M., Kulshreshta, R., et al. (2018) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2018.8462468 · Cited by 12
“We also examined the impact of training set size, by retraining the i-vector extractor on a large corpus of non-AMI wideband speech comprising SRE08, parts of Mixer 6 [14], and VoxCeleb [15].”
MethodChen, C., Wang, W., He, Y., & Han, J. (2019) · Digital Signal Processing · doi:10.1016/j.dsp.2019.03.008 · Cited by 12
Towards Real-Time Hidden Speaker Recognition by Means of Fully Homomorphic Encryption
Zuber, M., Carpov, S., & Sirdey, R. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-61078-4_23 · Cited by 12
“It was trained and tested on datasets called VoxCeleb1 [30] and VoxCeleb2 [16].”
MethodSun, L., Du, J., Zhang, X., Gao, T., Fang, X., & Lee, C. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9053317 · Cited by 12
“The corpora we used here include VoxCeleb1 [31] and VoxCeleb2 [32].”
MethodAdaptive Margin Circle Loss for Speaker Verification
Xiao, R., Miao, X., Wang, W., Zhang, P., Cai, B., & Luo, L. (2021) · Interspeech (USB) · doi:10.21437/interspeech.2021-1043 · Cited by 16
Zheng, N., Li, N., Wu, X., Meng, L., Kang, J., Wu, H., Weng, C., Su, D., & Meng, H. M. L. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746649 · Cited by 12
Neural PLDA Modeling for End-to-End Speaker Verification
Ramoji, S., Krishnan, P., & Ganapathy, S. (2020) · Interspeech (USB) · doi:10.21437/interspeech.2020-2699 · Cited by 13
HarperValleyBank: A Domain-Specific Spoken Dialog CorpusPreprint
Wu, M., Nafziger, J., Scodary, A., & Maas, A. L. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.13929 · Cited by 18
“0 There is significant recent work in representation learning for domain and task transfer via embedding models [14, 15, 16, 17], which can be trained without any supervision and reused for many downstream tasks like predicting speaker identity [18, 19] and commands [20, 21].”
BackgroundSpeaker De-identification System using Autoencoders and Adversarial TrainingPreprint
Espinoza-Cuadros, F., Perero-Codosero, J. M., Antón-Martín, J., & Gómez, L. A. H. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.04696 · Cited by 12
“For AAN training and testing we used VoxCeleb-1 [22], which contains approximately 330 hours of recordings from 1251 speakers.”
MethodFederated Learning of User Verification Models Without Sharing EmbeddingsPreprint
Hosseini, H., Park, H., Yun, S., Louizos, C., Soriaga, J. B., & Welling, M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.08776 · Cited by 22
“• We present the experimental results for voice, face and handwriting recognition using VoxCeleb (Nagrani et al., 2017), CelebA (Liu et al.”
BackgroundMethodInfluentialDEEPFAKER: A Unified Evaluation Platform for Facial Deepfake and Detection Models
Wang, L., Meng, X., Li, D., Zhang, X., Ji, S., & Guo, S. (2024) · ACM Transactions on Privacy and Security · doi:10.1145/3634914 · Cited by 11
Speaker Identity Preservation in Dysarthric Speech Reconstruction by Adversarial Speaker Adaptation
Wang, D., Liu, S., Wu, X., Lu, H. J., Sun, L., Liu, X., & Meng, H. M. L. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746680 · Cited by 11
Modified layer deep convolution neural network for text-independent speaker recognition
Karthikeyan, V., & Priyadharsini, S. S. (2022) · Journal of Experimental & Theoretical Artificial Intelligence · doi:10.1080/0952813x.2022.2092560 · Cited by 11
“Both feature extraction and classification audits can be performed at the same time in CNN (Mehri et al., 2017; Muckenhirn et al., 2018; Nagrani et al., 2017; Palaz et al., 2015).”
BackgroundLarge motion anime head animation using a cascade pose transform network
Zhang, J., Liu, C., Xian, K., & Cao, Z. (2022) · Pattern Recognition · doi:10.1016/j.patcog.2022.109181 · Cited by 13
StyleGAN Salon: Multi-View Latent Optimization for Pose-Invariant Hairstyle Transfer
Khwanmuang, S., Phongthawee, P., Sangkloy, P., & Suwajanakorn, S. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.00832 · Cited by 11
“Recently, HairFIT [8], a pose-invariant hairstyle transfer network, aligns the reference hair to match the input face pose using a optical flow-based hair alignment module, but requires training on multi-view dataset [21, 26].”
BackgroundMethodSLMIA-SR: Speaker-Level Membership Inference Attacks against Speaker Recognition Systems
Chen, G., Zhang, Y., & Song, F. (2024) · Network and Distributed System Security Symposium · doi:10.14722/ndss.2024.241323 · Cited by 28
“84% VGG-GE2E VGGVox40 [58], [56] verification GE2E [29] 4.”
One-Shot Neural Face Reenactment via Finding Directions in GAN’s Latent Space
Bounareli, S., Tzelepis, C., Argyriou, V., Patras, I., & Tzimiropoulos, G. (2024) · International Journal of Computer Vision · doi:10.1007/s11263-024-02018-6 · Cited by 17
“28 Qualitative results and comparisons for cross-subject reenact-ment on VoxCeleb1 (Nagrani et al., 2017) dataset.”
InfluentialAttribute-Guided Cross-Modal Interaction and Enhancement for Audio-Visual Matching
Wang, J., Zheng, A., Yan, Y., He, R., & Tang, J. (2024) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2024.3388949 · Cited by 14
“We perform audio-visual matching experiments on the Voxceleb [30] and VGGFace [31] datasets, which can achieve superior performance compared to state-of-the-art algorithms.”
MethodSpeaker Recognition with Deep Learning Approaches: A Review
Alenizi, A. S., & Al-Karawi, K. A. (2024) · Lecture notes in networks and systems · doi:10.1007/978-981-97-3289-0_39 · Cited by 11
Hypernetworks for Personalizing ASR to Atypical Speech
Müller-Eberstein, M., Yee, D., Yang, K., Mantena, G., & Lea, C. (2024) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00696 · Cited by 18
“The speaker verification model uses an Emphasized Channel Attention, Propagation and Aggregation Time Delay Neural Network (ECAPA-TDNN; Desplanques et al., 2020), and is trained on the VoxCeleb datasets (Nagrani et al., 2017; Chung et al., 2018; Nagrani et al., 2020).”
MethodChameleon: A Multimodal Learning Framework Robust to Missing Modalities
Liaqat, M. I., Nawaz, S., Zaheer, M. Z., Saeed, M. S., Sajjad, H., Schepper, T. D., Nandakumar, K., Khan, M. H., Gallo, I., & Schedl, M. (2025) · International Journal of Multimedia Information Retrieval · doi:10.1007/s13735-025-00370-y · Cited by 23
Hussain, S. M., Saritha, B., Reddy, B. S., Srikar, C., Suchitra, B., & Purnachandrarao, G. (2025) · International Conference on Electronics, AI and Computing (EAIC) · doi:10.1109/eaic66483.2025.11101389 · Cited by 12
“When using CNNs to examine speaker-specific spectral patterns, Nagrani et al. encountered difficulties capturing temporal relationships [10].”
Kim, I.-S., Kim, K. H., Kim, J., & Choi, C. (2019) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2019.8683332 · Cited by 11
“Figure 2 shows a comparison between results on the VoxCeleb dataset [9] from the best previous work (ResCNN-GAP) and our proposed method.”
MethodResultDIHARD II is Still Hard: Experimental Results and Discussions from the DKU-LENOVO Team
Lin, Q., Cai, W., Yang, L., Wang, J., Zhang, J., & Li, M. (2020) · The Speaker and Language Recognition Workshop (Odyssey) · doi:10.21437/odyssey.2020-15 · Cited by 11
Image Animations on Driving Videos with DeepFakes and Detecting DeepFakes Generated Animations
Malik, Y. S., Sabahat, N., & Moazzam, M. O. (2020) · IEEE International Multitopic Conference (INMIC) · doi:10.1109/inmic50486.2020.9318064 · Cited by 15
“Generative Adversarial Networks are capable of performing image animation and video retargeting [1,2,3].”
BackgroundMinimum Bayes Risk Training for End-to-End Speaker-Attributed ASR
Kanda, N., Meng, Z., Lu, L., Gaur, Y., Wang, X., Chen, Z., & Yoshioka, T. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9415062 · Cited by 18
Data Augmentation Methods for End-to-End Speech Recognition on Distant-Talk Scenarios
Tsunoo, E., Shibata, K., Narisetty, C. P., Kashiwagi, Y., & Watanabe, S. (2021) · Interspeech (USB) · doi:10.21437/interspeech.2021-958 · Cited by 13
CS-REP: Making Speaker Verification Networks Embracing Re-Parameterization
Zhang, R., Wei, J., Lu, W., Zhang, L., Ji, Y., Xu, J., & Lu, X. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746456 · Cited by 13
Federated Learning for Privacy Preserving On-Device Speaker Recognition
Woubie, A., & Bäckström, T. (2021) · doi:10.21437/spsc.2021-1 · Cited by 11
“The first method randomly selects the speech of other persons from Voxceleb dataset [10] as impostor speech for a given speaker.”
MethodAttention based gender and nationality information exploration for speaker identification
Tang, Y., Liu, C., Leng, Y., Zhao, W., Sun, J., Sun, C., Wang, R.-Y., Yuan, Q., Li, D., & Xu, H. (2022) · Digital Signal Processing · doi:10.1016/j.dsp.2022.103449 · Cited by 13
EDITnet: A Lightweight Network for Unsupervised Domain Adaptation in Speaker Verification
Li, J., Liu, W., & Lee, T. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-967 · Cited by 14
“The VoxCeleb 1&2 [25–27] is one of the most popular large-scale datasets for SV, with the speech data downloaded from YouTube 1 .”
BackgroundMotion and Appearance Adaptation for Cross-domain Motion Transfer
Xu, B., Wang, B., Deng, J., Tao, J., Ge, T., Jiang, Y., Li, W., & Duan, L. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19787-1_30 · Cited by 14
Composing General Audio Representation by Fusing Multilayer Features of a Pre-trained Model
Niizumi, D., Takeuchi, D., Ohishi, Y., Harada, N., & Kashino, K. (2022) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco55093.2022.9909674 · Cited by 15
Towards A Unified Conformer Structure: from ASR to ASV Task
Liao, D., Jiang, T., Wang, F., Li, L., & Hong, Q. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095433 · Cited by 14
“We conducted our proposed experiments on CN-Celeb [24, 25] and VoxCeleb [22, 23] respectively.”
MethodResultChauhan, N., Isshiki, T., & Li, D. (2023) · SN Computer Science · doi:10.1007/s42979-023-02056-w · Cited by 11
“NIST-2008 Bottleneck feature, i-vector – 5.86 [48] Voxceleb1 Score-level fusion, DNN, x-vector – 3.85 [48] Voxceleb1 Score-level fusion, i-vector – 5.39 [49] Voxceleb1 Automated pipeline, CNN 80.5 7.8”
MethodInfluentialControlVC: Zero-Shot Voice Conversion with Time-Varying Controls on Pitch and Speed
Chen, M., & Duan, Z. (2023) · Interspeech (USB) · doi:10.21437/interspeech.2023-1788 · Cited by 23
“The speaker encoder is pre-trained on a combination of VoxCeleb [22] and Librispeech [23] datasets with a total of 3,549 speakers using GE2E loss (3).”
MethodA multi-purpose audio-visual corpus for multi-modal Persian speech recognition: The Arman-AV dataset
Peymanfard, J., Heydarian, S., Lashini, A., Zeinali, H., Mohammadi, M. R., & Mozayani, N. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.121648 · Cited by 20
Analysis of Oral Exams With Speaker Diarization and Speech Emotion Recognition: A Case Study
Beccaro, W., Ramírez, M. A., Liaw, W., & Guimarães, H. R. (2023) · IEEE Transactions on Education · doi:10.1109/te.2023.3321155 · Cited by 12
“…files were preprocessed with a pretrained voice activity detector (VAD) [40] followed by an algorithm to extract speaker x-vector embeddings [41] combined with a pretrained time delay neural network (TDNN), which achieved an equal error rate (EER) of 3.2% on the clean VoxCeleb1 test set [42], [43].”
MethodInfluentialAudio-Visual Speaker Verification via Joint Cross-Attention
Praveen, R. G., & Alam, J. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-48312-7_2 · Cited by 14
Randomized Quantization: A Generic Augmentation for Data Agnostic Self-supervised Learning
Wu, H., Lei, C., Sun, X., Wang, P., Chen, Q., Cheng, K.-T. T., Lin, S. C.-F., & Wu, Z. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.01494 · Cited by 22
“Task #Classes Data size Avg duration (s) NSynth (NS) [27] Musical instrument classification 11 305,979 4.0 UrbanSound8K (US8K) [60] Urban sound classification 10 8,732 4.0 VoxCeleb1 (VC1) [53] Speaker identification Comparison with weakly-augmented SSL baselines.”
BackgroundInfluentialLi, Z., Mak, M., & Meng, H. M. L. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10447795 · Cited by 14
“Specifically, it performs comparably to full fine-tuning with only 20.2M parameters for the prompt vectors and adapters and 14.7M parameters for the ECAPA-TDNN.”
MethodSholokhov, A., Kinnunen, T., Vestman, V., & Lee, K. A. (2019) · Computer Speech & Language · doi:10.1016/j.csl.2019.101024 · Cited by 11
“As a representative snapshot of the current ASV technology and evaluation databases, our proof-of-concept experiments include two widely-used ASV methods based on i-vector [22] and x-vector [23] embeddings, evaluated on the combined VoxCeleb1 [24] and VoxCeleb2 [25] corpora.”
MethodPham, T. M., Li, Z., & Whitehill, J. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9053258 · Cited by 11
“Using this approach, VoxCeleb1 and VoxCeleb2 were created which contain 1,251 and 6,112 unique speakers, respectively.”
BackgroundAn Effective Speaker Recognition Method Based on Joint Identification and Verification Supervisions
Liu, Y., Song, Y., Jiang, Y., McLoughlin, I. V., Liu, L., & Dai, L.-R. (2020) · Interspeech (USB) · doi:10.21437/interspeech.2020-1922 · Cited by 11
“To evaluate the effectiveness of the proposed method, extensive experiments have been conducted on the Voxceleb1 benchmark [18].”
MethodLearning spectro-temporal representations of complex sounds with parameterized neural networks
Riad, R., Karadayi, J., Bachoud‐Lévi, A., & Dupoux, E. (2021) · The Journal of the Acoustical Society of America · doi:10.1121/10.0005482 · Cited by 13
“We also reported the baseline performance of the I-vector system trained on VoxCeleb1 combined with Probabilistic Linear Discriminant Analysis (PLDA) (Dehak et al., 2010).”
MethodKataria, S., Villalba, J., Żelasko, P., Moro-Velázquez, L., & Dehak, N. (2021) · Interspeech (USB) · doi:10.21437/interspeech.2021-1502 · Cited by 11
“Data augmentation is not performed during CycleGAN training and VoxCeleb2 is downsam-pled a priori to 8 kHz to match the telephone sampling rate.”
InfluentialHybrid Speech and Text Analysis Methods for Speaker Change Detection
Anidjar, O. H., Lapidot, I., Hajaj, C., Dvir, A. Z., & Gilad, I. (2021) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2021.3093817 · Cited by 11
BIOMEX-DB: A Cognitive Audiovisual Dataset for Unimodal and Multimodal Biometric Systems
Moreno-Rodríguez, J. C., Atenco-Vázquez, J. C., Ramírez-Cortés, J. M., Arechiga, R. O., Gómez‐Gil, P., & Fonseca-Delgado, R. (2021) · IEEE Access · doi:10.1109/access.2021.3100035 · Cited by 14
“So far, this project has made public two datasets: VoxCeleb1 [20] and VoxCeleb2 [21], both originally meant to perform speaker recognition experiments.”
Global–Local Self-Attention Based Transformer for Speaker Verification
Xie, F., Zhang, D., & Liu, C. (2022) · Applied Sciences · doi:10.3390/app121910154 · Cited by 14
“Performance comparison with the VoxCeleb1 test set.”
MethodInfluentialExact Prosody Cloning in Zero-Shot Multispeaker Text-to-Speech
Lux, F., Koch, J., & Vu, N. T. (2023) · 2022 IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt54892.2023.10022433 · Cited by 11
“Both are trained on Voxceleb 1 and 2 [23, 24, 25] using the SpeechBrain toolkit [26].”
MethodDeep Multi-biometric Fusion for Audio-Visual User Re-Identification and Verification
Marras, M., Marín-Reyes, P. A., Lorenzo-Navarro, J., Castrillón-Santana, M., & Fenu, G. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-40014-9_7 · Cited by 14
Meeting Transcription Using Virtual Microphone ArraysPreprint
Yoshioka, T., Chen, Z., Dimitriadis, D., Hinthorn, W., Huang, X., Stolcke, A., & Zeng, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.02545 · Cited by 11
“We trained a ResNet-style embedding extraction network [19] on the VoxCeleb corpus [20] to generate 128-dimensional dvectors.”
MethodSpeaker Diarization as a Fully Online Learning Problem in MiniVoxPreprint
Lin, B., & Zhang, X. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.04376 · Cited by 11
ASVtorch toolkit: Speaker verification with deep neural networks
Lee, K. A., Vestman, V., & Kinnunen, T. (2021) · SoftwareX · doi:10.1016/j.softx.2021.100697 · Cited by 13
“177 The VoxCeleb1 and VoxCeleb2 datasets used for VoxCeleb evaluation 178 were collected from YouTube by the authors of [14] and [15].”
MethodInfluentialPractical Attacks on Voice Spoofing CountermeasuresPreprint
Kassis, A., & Hengartner, U. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.14642 · Cited by 18
“The target systems, hence, are ASV-CMs, where the ASV may be either a GMM i-vector (which is identical to asvs from earlier), or a GMM x-vector, trained on a different dataset (the VoxCeleb dataset [48]).”
BackgroundCai, D., & Li, M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2109.02853 · Cited by 11
“For evaluation, the development set and test set of Voxceleb 1 are used [11].”
MethodIn Search of Strong Embedding Extractors for Speaker Diarisation
Jung, J.-W., Heo, H.-S., Lee, B., Huh, J., Brown, A., Kwon, Y., Watanabe, S., & Chung, J. S. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096449 · Cited by 17
“We adopt the development sets of the VoxCeleb1&2 datasets [17,36] as the training set.”
MethodSelf-citationGeneralizing Speaker Verification for Spoof Awareness in the Embedding Space
Liu, X., Sahidullah, M., Lee, K. A., & Kinnunen, T. (2024) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2024.3358056 · Cited by 22
“2(b), the two embed-dings that correspond to the enrolment and test are concatenated as inputs of the backend network.”
MethodCavalcanti, J. C., da Silva, R. R., Eriksson, A., & Barbosa, P. A. (2024) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2024.1287877 · Cited by 14
“• All of these audio samples were processed by the ECAPA-TDNN model (Desplanques et al., 2020) provided by the Speechbrain project (Desplanques et al., 2021), which was pre-trained using the training data from the Voxceleb1 (Nagrani et al., 2017) and Voxceleb2 (Chung et al., 2018) databases, obtaining a 192-dimensiona…”
MethodCosine Scoring With Uncertainty for Neural Speaker Embedding
Wang, Q., & Lee, K. A. (2024) · IEEE Signal Processing Letters · doi:10.1109/lsp.2024.3375080 · Cited by 15
“We evaluate the proposed methods on both the Vox-Celeb1 [34] and the Speaker in the Wild (SITW) core-core eval sets [35].”
Adaptive Super Resolution for One-Shot Talking-Head Generation
Song, L., Liu, P., Yin, G., & Xu, C. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10446837 · Cited by 16
“Pretrained on large-scale datasets [10, 11], these methods transform pixels based on the flow-guided warp or Jacobin matrix [5] to obtain the driven image.”
MethodMobile Virtual Assistant for Multi-Modal Depression-Level Stratification
Wu, E. H., Gao, T., Chung, C., Chen, C., Tsai, C.-F., & Yeh, S. (2024) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2024.3451114 · Cited by 11
“FAb-Net has been trained on the VoxCeleb1 [63] and VoxCeleb2 [64] datasets involving interview videos containing over 1 million utterancesforaround7000speakers.”
Kim, M., & Jang, G. (2024) · Applied Sciences · doi:10.3390/app14188138 · Cited by 11
“Both of these models were trained on the augmented VoxCeleb 1 and 2 datasets [40].”
MethodInfluentialTowards Real-world Video Face Restoration: A New Benchmark
Chen, Z., He, J., Lin, X., Qiao, Y., & Dong, C. (2024) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw63382.2024.00600 · Cited by 19
Fatehi, K., Torres, M. T., & Küçükyılmaz, A. (2024) · Speech Communication · doi:10.1016/j.specom.2024.103151 · Cited by 17
Jha, K., Jain, A., & Srivastava, S. (2025) · Technology and Health Care · doi:10.1177/09287329251363424 · Cited by 10
“Dataset1 Description: VoxCeleb1 31 is an extensive audio-visual dataset comprising human voice samples obtained from YouTube videos.”
InfluentialSingingHead: A Large-Scale 4D Dataset for Singing Head Animation
Wu, S., Li, Y., Zhang, W., Jia, J., Zhu, Y., Yan, Y., Zhai, G., & Yang, X. (2025) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2025.3623560 · Cited by 24
“Recently, impressive progress has been made in speech-driven talking head animation [12, 30, 53, 58] thanks to the release of a variety of talking head datasets [8, 13, 36, 50, 59].”
BackgroundMethodBayesian x-vector: Bayesian Neural Network based x-vector System for Speaker Verification
Xu, L., Zhong, J., Yu, J., Hu, S., Wu, X., Liu, X., & Meng, H. M. L. (2020) · The Speaker and Language Recognition Workshop (Odyssey) · doi:10.21437/odyssey.2020-51 · Cited by 10
Acoustic and Linguistic Analyses to Assess Early-Onset and Genetic Alzheimer’s Disease
Pérez-Toro, P. A., Vásquez-Correa, J. C., Arias‐Vergara, T., Klumpp, P., Castrillón, M. S., Roldan-Lopez, M. E., Aguillon, D. F., Hincapié, L., Tobón, C., Bocklet, T., et al. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9414009 · Cited by 14
“The pretrained model uses augmented Vox-Celeb [17].”
MethodInfluentialMemory Layers with Multi-Head Attention Mechanisms for Text-Dependent Speaker Verification
Mingote, V., Miguel, A., Ortega, A., & Lleida, E. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9414859 · Cited by 11
“In the second set of experiments using DeepMine-part I, the baseline to compare is based on x-vectors presented by organizers of SdSV Challenge [20] which apart from training with the mention DeepMine database, this baseline was trained using the popular VoxCeleb 1 and 2 databases [23, 24].”
MethodFsft-Net: Face Transfer Video Generation With Few-Shot Views
Song, L., Yin, G., Liu, B., Zhang, Y., & Yu, N. (2021) · Proceedings - International Conference on Image Processing · doi:10.1109/icip42928.2021.9506512 · Cited by 14
“To demonstrate the effectiveness of our proposed method, we conduct extensive experiments on large-scale video dataset VoxCeleb[9].”
MethodStudying Squeeze-and-Excitation Used in CNN for Speaker Verification
Rouvier, M., & Bousquet, P.-M. (2021) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru51503.2021.9687936 · Cited by 20
Multisv: Dataset for Far-Field Multi-Channel Speaker Verification
Mošner, L., Plchot, O., Burget, L., & Černocký, J. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746833 · Cited by 13
Lightweight Speaker Recognition in Poincaré Spaces
Lee, J., Kim, S.-B., Kang, S., & Oh, T.-H. (2021) · IEEE Signal Processing Letters · doi:10.1109/lsp.2021.3129695 · Cited by 10
Gu, B., & Guo, W. (2021) · IEEE Signal Processing Letters · doi:10.1109/lsp.2021.3136141 · Cited by 11
A deep learning approach for text-independent speaker recognition with short utterances
Chakroun, R., & Frikha, M. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-14942-9 · Cited by 10
Sun, K., Xia, C., Xu, S., & Zhang, X. (2023) · Network and Distributed System Security Symposium · doi:10.14722/ndss.2023.24077 · Cited by 10
“We first embed the frame level feature by using TDNN blocks [29], and then apply an attentive statistics pooling layer [30] to convert the variable length frame-level features into a fixed-dimensional vector, referred to as deep speaker embedding in text-independent and variable-duration scenarios [31].”
MethodLiu, X., Li, X., & Serrà, J. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096478 · Cited by 15
“Note that, in the following, SID models without ‘FN’ in their names do not use feature normalization, and that models without ‘aug’ in their names are trained on VoxCeleb1,2 [22,23] with 7,205 speakers without data augmentation.”
MethodDifferentially Private Adversarial Auto-Encoder to Protect Gender in Voice Biometrics
Chouchane, O., Panariello, M., Zari, O., Kerenciler, I., Chihaoui, I., Todisco, M., & Önen, M. (2023) · ACM Workshop on Information Hiding and Multimedia Security (IH&MMSec) · doi:10.1145/3577163.3595102 · Cited by 15
“The external sex classifier is trained using a subset of the VoxCeleb1 development partition (61616 segments per class).”
MethodInfluentialEnd-to-end speaker identification research based on multi-scale SincNet and CGAN
Wei, G., Zhang, Y., Min, H., & Xu, Y. (2023) · Neural Computing and Applications · doi:10.1007/s00521-023-08906-1 · Cited by 10
Fake the Real: Backdoor Attack on Deep Speech Classification via Voice Conversion
Ye, Z., Mao, T., Dong, L., & Yan, D. (2023) · Interspeech (USB) · doi:10.21437/interspeech.2023-733 · Cited by 23
“Additionally, we selected two speech datasets for speaker recognition: TIMIT [26] and VoxCeleb1 [27].”
A Lightweight CNN-Conformer Model for Automatic Speaker Verification
Wang, H., Lin, X., & Zhang, J. (2023) · IEEE Signal Processing Letters · doi:10.1109/lsp.2023.3342714 · Cited by 12
Learning an Interpretable Stylized Subspace for 3D-Aware Animatable Artforms
Zheng, C., Liu, B., Xu, X., Zhang, H., & He, S. (2024) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2024.3364162 · Cited by 11
Efficient Adapter Tuning of Pre-Trained Speech Models for Automatic Speaker Verification
Sang, M., & Hansen, J. H. L. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10446686 · Cited by 18
“The speaker verification systems are trained and evaluated on the VoxCeleb1 [30] dataset, which contains 148642 utterances from 1211 speakers in the development set and 4874 utterances from 40 speakers in the test set.”
MethodSpontts: Modeling and Transferring Spontaneous Style for TTS
Li, H., Zhu, X., Xue, L., Song, Y., Chen, Y., & Xie, L. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10445828 · Cited by 19
“We adopt ECAPA-TDNN [24] to extract speaker embedding, which is trained on voxceleb with 351 hours and 1251 speakers [25].”
Method
Metadata
- Authors
- Nagrani, A., Chung, J. S., & Zisserman, A.
- Venue
- Interspeech 2017
- Publisher
- ISCA
- Published
- 2017-08-20
- Pages
- 2616-2620
- ISSN
- 2308-457X (ISSN-L)
- OpenAlex
- W2726515241
- Cited by
- 2,210
- References
- 4
- Citations by year
- 2026: 144 · 2025: 280 · 2024: 313 · 2023: 361 · 2022: 330 · 2021: 319 · 2020: 265 · 2019: 151 · 2018: 47
- FWCI
- 56.51
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Interspeech (USB) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Engineering and Physical Sciences Research Council (EP/M013774/1)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Physical Sciences › Computer Science › Signal Processing › Speech and Audio Processing
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face recognition and analysis
- Keywords
- speaker identification, speaker verification, VoxCeleb, text-independent speaker recognition, automated dataset generation, two-stream CNN, face recognition, YouTube videos
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.