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 801–900 of 1,008
Choi, S., Chung, S., Lee, S., Han, S., Kang, T., Seo, J., Kwak, I., & Oh, S. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10448221 · Cited by 11
“Model performance is evaluated on the VoxCeleb1 test set [19], comprising 1,251 speakers and 153,516 utterances.”
MethodZhong, Q., Dai, R., Zhang, H., Zhu, Y., & Zhou, G. (2021) · EURASIP Journal on Advances in Signal Processing · doi:10.1186/s13634-021-00762-2 · Cited by 10
“6, when the CASP was used as the aggregation layer and the ACLL, AAM, and AM were used as the loss function, the average EER was 1.76%, 1.91%, and 1.98% on VoxCeleb1; 1.91%, 2.04%, and 2.06% on VoxCeleb1-E; and 3.24%, 3.38%, and 3.44% on VoxCeleb1-H, respectively.”
MethodInfluentialA Hybrid GRU-CNN Feature Extraction Technique for Speaker Identification
Shihab, M. S. H., Aditya, S., Setu, J. H., Imtiaz-Ud-Din, K. M., & Efat, M. I. A. (2020) · International Conference on Computer and Information Technology (ICCIT) · doi:10.1109/iccit51783.2020.9392734 · Cited by 10
“The speakers span a wide run of distinctive ethnicity’s, emphasizes, callings and ages [24].”
BackgroundMethodStreaming Multi-Talker Speech Recognition with Joint Speaker Identification
Lu, L., Kanda, N., Li, J., & Gong, Y. (2021) · Interspeech (USB) · doi:10.21437/interspeech.2021-207 · Cited by 14
Adversarial Attack and Defense on Deep Neural Network-Based Voice Processing Systems: An Overview
Chen, X., Li, S., & Huang, H. (2021) · Applied Sciences · doi:10.3390/app11188450 · Cited by 12
Group-Level Focus of Visual Attention for Improved Next Speaker Prediction
Birmingham, C., Stefanov, K., & Matarić, M. J. (2021) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3474085.3479213 · Cited by 10
“The output of the final convolutional layer from the PerfectMatch model that had been pre-trained on VoxCeleb [12] was directly used to produce 512 𝐷 audio and 512 𝐷 video features for each frame.”
MethodKang, W. H., Alam, J., & Fathan, A. (2021) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru51503.2021.9687877 · Cited by 10
An Analytic Study on Clustering-Based Pseudo-labels for Self-supervised Deep Speaker Verification
Kang, W. H., Alam, J., & Fathan, A. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20980-2_29 · Cited by 11
Distance-Based Weight Transfer for Fine-Tuning From Near-Field to Far-Field Speaker Verification
Zhang, L., Wang, Q., Wang, H., Li, Y., Rao, W., Wang, Y., & Xie, L. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096790 · Cited by 11
“In recent years, deep learning has shown remarkable success in SV tasks, but current methods often rely on a large amount of labeled training speech [2, 3].”
BackgroundFathan, A., & Alam, J. (2023) · International Workshop on Biometrics and Forensics (IWBF) · doi:10.1109/iwbf57495.2023.10157651 · Cited by 10
“The evaluation was performed according to the original VoxCeleb1 trials list [45], which consists of 4,874 utterances spoken by 40 speakers.”
MethodLei, L., & She, K. (2018) · Entropy · doi:10.3390/e20080600 · Cited by 10
“In this paper, the TIMIT [25] and Voxceleb [26] speech corpus are used for experiments.”
MethodMarras, M., Marín-Reyes, P. A., Lorenzo-Navarro, J., Castrillón-Santana, M., & Fenu, G. (2019) · International Conference on Pattern Recognition Applications and Methods · doi:10.5220/0007690902550265 · Cited by 12
“The recent widespread of deep learning in different areas (e.g., Boratto et al. 2016, Nagrani et al. 2017) has motivated the use of the neural networks as feature extractors combined with classifiers, as proposed in Wang et al. 2018b.”
BackgroundMethodInfluentialFernando, T., Sridharan, S., McLaren, M., Priyasad, D., Denman, S., & Fookes, C. (2020) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2020.2982297 · Cited by 11
“Similar to [55] inputs are normalised to a range 0-1, and no other speech-specific preprocessing is performed.”
BackgroundContext and Uncertainty Modeling for Online Speaker Change Detection
Aronowitz, H., & Zhu, W. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9053280 · Cited by 11
“Training data is a combination of GALE training data, with some additional IBM internal wideband resources, and data extracted from the OpenSLR [30] and VoxCeleb1 [23].”
MethodAn Adaptive X-Vector Model for Text-Independent Speaker Verification
Gu, B., Guo, W., Ding, F., Ling, Z.-H., & Du, J. (2020) · Interspeech (USB) · doi:10.21437/interspeech.2020-1071 · Cited by 10
“If not specified, all of the setups are the same as the baseline system.”
BackgroundA simulation study on optimal scores for speaker recognition
Wang, D. (2020) · EURASIP Journal on Audio Speech and Music Processing · doi:10.1186/s13636-020-00183-3 · Cited by 13
VGGSound: A Large-scale Audio-Visual DatasetPreprint
Chen, H., Xie, W., Vedaldi, A., & Zisserman, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.14368 · Cited by 39
“In contrast to these manually curated datasets, recent papers have demonstrated the possibility of collecting high-quality human speech datasets in an automated and scalable manner by using computer vision algorithms [11, 12, 13].”
ResultSelf-citationQASR: QCRI Aljazeera Speech Resource -- A Large Scale Annotated Arabic Speech CorpusPreprint
Mubarak, H., Hussein, A., Chowdhury, S. A., & Ali, A. M. A. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.13000 · Cited by 18
“The model is trained on Voxceleb1 (Nagrani et al., 2017) development set (containing 1 , 211 speakers and ≈ 147 K utterances).”
MethodA Survey on Audio Synthesis and Audio-Visual Multimodal ProcessingPreprint
Shi, Z. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.00443 · Cited by 22
“VoxCeleb The VoxCeleb dataset [63] is a large scale text-independent speaker identification dataset, which contains over 100000 utterances for 1251 celebrities.”
BackgroundInfluentialTopological Data Analysis for Speech Processing
Tulchinskii, E., Kuznetsov, K., Kushnareva, L., Cherniavskii, D., Barannikov, S., Piontkovskaya, I., Nikolenko, S., & Burnaev, E. V. (2023) · Interspeech (USB) · doi:10.21437/interspeech.2023-1861 · Cited by 12
Attention and DCT Based Global Context Modeling for Text-Independent Speaker Recognition
Xia, W., & Hansen, J. H. L. (2023) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2023.3284521 · Cited by 10
Rethinking one-shot face reenactment: A spatial–temporal reconstruction view
Chen, Y., & Xiong, S. (2023) · Knowledge-Based Systems · doi:10.1016/j.knosys.2023.110786 · Cited by 9
Choi, J., Yang, J.-Y., & Chang, J. (2024) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2024.3463491 · Cited by 10
A comprehensive study of audio profiling: Methods, applications, challenges, and future directions
Pudasaini, A., Al-Hawawreh, M. S., Bouadjenek, M. R., Hacid, H., & Aryal, S. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.130334 · Cited by 9
“For instance, the TIMIT dataset has 192 female and 438 male speakers, while the VoxCeleb dataset [252], though more balanced, still has a skew with 55% male and 45% female speakers.”
MethodMaciejewski, M., Sell, G., Fujita, Y., Garcia, L. P., Watanabe, S., & Khudanpur, S. P. (2019) · IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) · doi:10.1109/waspaa.2019.8937153 · Cited by 9
“The models were trained using the VoxCeleb [25] and VoxCeleb2 [26] corpora augmented with MUSAN [21] and reverberated with the simulated room impulse responses described in [27].”
MethodEfficient and Privacy-Preserving Speaker Recognition for Cybertwin-Driven 6G
Li, Q., & Lin, X. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3097266 · Cited by 9
Privacy-Protecting Techniques for Behavioral Data: A Survey
Hanisch, S., Arias-Cabarcos, P., Parra‐Arnau, J., & Strufe, T. (2021) · Repository KITopen (Karlsruhe Institute of Technology) · doi:10.5445/ir/1000139989 · Cited by 10
Dual-Path Modeling With Memory Embedding Model for Continuous Speech Separation
Li, C., Chen, Z., & Qian, Y. (2022) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2022.3165712 · Cited by 9
Voice-quality Features for Deep Neural Network Based Speaker Verification Systems
Woubie, A., Koivisto, L., & Bäckström, T. (2021) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco54536.2021.9616242 · Cited by 9
Domain Agnostic Few-shot Learning for Speaker Verification
Yang, S.-H., Das, D., Cho, J., Park, H. W., & Yun, S. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-940 · Cited by 9
Multi-Voice Singing Synthesis From Lyrics
Resna, S., & Rajan, R. (2022) · Circuits Systems and Signal Processing · doi:10.1007/s00034-022-02122-3 · Cited by 11
“LibriSpeech [30], VoxCeleb1 [26] andVoxCeleb2[27] are used for training and validation of theWaveNet encoder in TTSmodule.”
MethodTraining speaker recognition systems with limited data
Vaessen, N., & van Leeuwen, D. A. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-135 · Cited by 10
“There is no overlap between VoxCeleb1 [42] and Voxceleb2 [24].”
BackgroundMethodSAKA: an intelligent platform for semi-automated knowledge graph construction and application
Zhang, H., Wang, X., Pan, J., & Wang, H. (2023) · Service Oriented Computing and Applications · doi:10.1007/s11761-023-00371-x · Cited by 11
“The VoxCeleb is a large-scale dataset for speaker identification.”
MethodA Fused Speech Enhancement Framework for Robust Speaker Verification
Wu, Y., Li, T., Zhao, J., Wang, Q., & Xu, J. (2023) · IEEE Signal Processing Letters · doi:10.1109/lsp.2023.3290832 · Cited by 12
“(cid:2) Experiments conducted on noise-simulated VoxCeleb1 [28] dataset shows that both the FSEF and the MSACAN can improve the performance of RSV compared to previous approaches.”
MethodInfluentialVoiceLDM: Text-to-Speech with Environmental Context
Lee, Y., Yeon, I., Nam, J., & Chung, J. S. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10448268 · Cited by 39
“We use the following publicly available real-world audio datasets for training: AudioSet [22], the English subset of the CommonVoice 13.0 corpus [23], VoxCeleb1 [24], and DEMAND [25].”
InfluentialSelf-citationContinuous Piecewise-Affine Based Motion Model for Image Animation
Wang, H., Liu, F., Zhou, Q., Yi, R., Tan, X., & Ma, L. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i6.28351 · Cited by 16
Yin, S., Chen, B., Wang, S., & Yan, Y. (2024) · DCC · doi:10.1109/dcc58796.2024.00019 · Cited by 9
Fang, Z., He, L. J., Li, L., & Hu, Y. (2024) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2024.3407527 · Cited by 10
Ghorbani, S., & Hansen, J. H. L. (2024) · The Journal of the Acoustical Society of America · doi:10.1121/10.0026235 · Cited by 18
Cross-Attention Fusion of Visual and Geometric Features for Large-Vocabulary Arabic Lipreading
Daou, S., Hamadou, A. B., Rekik, A., & Kallel, A. (2025) · Technologies · doi:10.3390/technologies13010026 · Cited by 12
Codec-Superb @ SLT 2024: A Lightweight Benchmark For Neural Audio Codec Models
Wu, H., Chen, X., Lin, Y., Chang, K., Du, J., Lu, K.-H., Liu, A. H., Chung, H.-L., Wu, Y.-K., Yang, D., et al. (2024) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt61566.2024.10832364 · Cited by 24
“Librispeech [31] diverse speaker, read audiobooks ✓ ✓ VoxCeleb1 [32] diverse speaker, celebrities on YouTube ✓ ✓ QUESST [33] multi-lingual, low resource language ✓ VoxLingua107 Top 10 [34] multi-lingual, YouTube content ✓ Fluent Speech Commands [35] spoken keyword commands ✓ Audio SNIPS [36] spoken…”
3D Convolutional Neural Networks Based Speaker Identification and Authentication
Liao, J., Wang, S., Zhang, X., & Liu, G. (2018) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2018.8451204 · Cited by 9
“Since most public speaker identification/authentication datasets are designed using the face/voice feature (such as VoxCeleb [20], etc.), the lip region in each frame is of too low resolution (less than 50 × 30 pixels) to extract useful lip features.”
BackgroundQuality Control in Remote Speech Data Collection
Poorjam, A. H., Little, M. A., Jensen, J. R., & Christensen, M. G. (2019) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2019.2904212 · Cited by 9
Speech-Driven Facial Animation Using Polynomial Fusion of Features
Kefalas, T., Vougioukas, K., Panagakis, Y., Petridis, S., Kossaifi, J., & Pantić, M. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9054469 · Cited by 9
“[2] which uses concatenation as its fusion layer, and the Speech2Vid model [3], which is trained on the Voxceleb [20] and LRW [21] datasets.”
MethodDeep Neural Networks with Batch Speaker Normalization for Intoxicated Speech Detection
Wang, W., Wu, H., & Li, M. (2019) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipaasc47483.2019.9023074 · Cited by 9
“Therefore, for baseline 3, we extract the i-vector of each utterance in testing data with a pre trained GMM-UBM based speaker recoguition system, which is trained on the voxceleb[29] with kaldi toolkit[30].”
MethodSpeaker Recognition Using SincNet and X-Vector Fusion
Tripathi, M., Singh, D., & Susan, S. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-61401-0_24 · Cited by 9
Intermediary Fuzzification in Speech Emotion Recognition
Assunção, G., & Menezes, P. J. (2020) · IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) · doi:10.1109/fuzz48607.2020.9177699 · Cited by 12
Xia, W., & Hansen, J. H. L. (2020) · Interspeech (USB) · doi:10.21437/interspeech.2020-1845 · Cited by 9
Adversarial Separation and Adaptation Network for Far-Field Speaker Verification
Yi, L., & Mak, M.-W. (2020) · Interspeech · doi:10.21437/interspeech.2020-2372 · Cited by 8
“The combination of VoxCeleb1, VoxCeleb2, and their augmented sets were used to train an x-vector extractor.”
BackgroundInfluentialEnd-To-End Lip Synchronisation Based on Pattern Classification
Kim, Y. J., Heo, H. S., Chung, S.-W., & Lee, B. (2021) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt48900.2021.9383616 · Cited by 14
“An off-sync video is not only unnatural to the human eye [1], but also causes errors in audio-visual speech enhancement [2, 3] and in making datasets for speaker recognition [4] and lip reading [5].”
BackgroundSingle Source One Shot Reenactment using Weighted Motion from Paired Feature Points
Tripathy, S., Kannala, J., & Rahtu, E. (2022) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv51458.2022.00218 · Cited by 13
Luo, C., Guo, X. Y., Deng, A., Xu, W. H., Zhao, J., & Kang, W. (2021) · IEEE International Joint Conference on Biometrics (IJCB) · doi:10.1109/ijcb52358.2021.9484331 · Cited by 9
Generating identities with mixture models for speaker anonymization
Turner, H., Lovisotto, G., & Martinović, I. (2021) · Computer Speech & Language · doi:10.1016/j.csl.2021.101318 · Cited by 19
Unifying Cosine and PLDA Back-ends for Speaker Verification
Peng, Z., He, X., Ding, K., Lee, T., & Wan, G. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-10021 · Cited by 9
HiFi-SVC: Fast High Fidelity Cross-Domain Singing Voice Conversion
Zhou, Y., & Lu, X. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746812 · Cited by 15
Duality Temporal-Channel-Frequency Attention Enhanced Speaker Representation Learning
Zhang, L., Wang, Q., & Xie, L. (2021) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru51503.2021.9688243 · Cited by 16
A study of bias mitigation strategies for speaker recognition
Peri, R., Somandepalli, K., & Narayanan, S. S. (2022) · Computer Speech & Language · doi:10.1016/j.csl.2022.101481 · Cited by 22
“dataset [104] as an external corpus (different from the MCV corpus) to evaluate the generalizability of the described methods on out-of-domain data.”
MethodImproving Speech Emotion Recognition via Fine-tuning ASR with Speaker Information
Thang, T. B., Nguyễn, T. L., Dang, D. S., Le, N. M., & Hai, V. (2022) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.23919/apsipaasc55919.2022.9980214 · Cited by 12
“1) ASR and ASV pre-training: The pre-trained ASR is trained on 960 hours Librispeech [33] dataset while the pretrained ASV uses Voxceleb1 [34] and Voxceleb2 [35] dataset with 7205 speakers and 1.”
MethodInfluentialDeepFake and its Enabling Techniques: A Review
Brooks, R., Yuan, Y., Liu, Y., & Chen, H. (2022) · APSIPA Transactions on Signal and Information Processing · doi:10.1561/116.00000024 · Cited by 12
UltraBD: Backdoor Attack against Automatic Speaker Verification Systems via Adversarial Ultrasound
Ze, J., Li, X., Cheng, Y., Ji, X., & Xu, W. (2023) · IEEE International Conference on Parallel and Distributed Systems (ICPADS) · doi:10.1109/icpads56603.2022.00033 · Cited by 16
“We adopt two widely-used human speech datasets to examine the performance of UltraBD , i.e., VoxCeleb [7] and LibriSpeech [8].”
MethodA Speech Representation Anonymization Framework via Selective Noise Perturbation
Tran, M., & Soleymani, M. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095173 · Cited by 10
“We use the VoxCeleb1 dataset [16] for training the SID and ASV models.”
MethodIncorporating Uncertainty from Speaker Embedding Estimation to Speaker Verification
Wang, Q., Lee, K. A., & Liu, T. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10097019 · Cited by 15
Sha, B., Xu, L., Wu, Z., Shan, Y., & Meng, H. M. L. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10446066 · Cited by 23
“Specifically, we train an ECAPA-TDNN [26] model on a large-scale combined dataset, which consists of the VoxCeleb1 and Vox-Celeb2 [27, 28], and the CN-Celeb1 and CN-Celeb2 [29, 30], total 3200+ hours of speech from 10k+ speakers.”
MethodCapturing Suprasegmental Features of a Voice with RNNs for Improved Speaker Clustering
Stadelmann, T., Glinski-Haefeli, S., Gerber, P., & Dürr, O. (2018) · Lecture notes in computer science · doi:10.1007/978-3-319-99978-4_26 · Cited by 9
“Despite speaker recognition progress in more challenging environments like meeting diarization [1] or less constrained speaker identification [30], speaker clustering of even just 40 speakers from TIMIT only recently gave reasonable results [24], while attempting to cluster all 630 speakers failed altogether [41].”
MethodResultI-Vector Transformation Using K-Nearest Neighbors for Speaker Verification
Khan, U. M., India, M., & Hernando, J. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9053504 · Cited by 9
“The experiments were performed on VoxCeleb-1 database [21, 22].”
MethodA Principle Solution for Enroll-Test Mismatch in Speaker Recognition
Li, L., Wang, D., Kang, J., Wang, R., Wu, J., Gao, Z., & Chen, X. (2022) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2022.3140558 · Cited by 9
Person Recognition with HGR Maximal Correlation on Multimodal Data
Liang, Y., Ma, F., Li, Y., & Huang, S. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9413164 · Cited by 9
“These methods can be divided into classification-based methods which directly learn identities in training data [3]–[6] and similarity-based learning methods which learn discriminative embeddings for different persons [7]–[11].”
BackgroundMethodInfluentialZhu, W., Kong, T., Lu, S., Li, J., Zhang, D., Deng, F., Wang, X., Yang, S., & Liu, J. (2021) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru51503.2021.9688017 · Cited by 9
Lepage, T., & Dehak, R. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-802 · Cited by 9
The DKU-OPPO System for the 2022 Spoofing-Aware Speaker Verification Challenge
Wang, X., Qin, X., Wang, Y., Xu, Y., & Li, M. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-11190 · Cited by 20
Fine-Grained Early Frequency Attention for Deep Speaker Representation Learning
Hajavi, A., & Etemad, A. (2023) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2023.3240113 · Cited by 11
“[8] VGG Softmax – 1024 TAP VoxCeleb1 7.”
BackgroundMethodInfluential$M^3$T: Multi-Modal Continuous Valence-Arousal Estimation in the WildPreprint
Zhang, Y., Huang, R., Zeng, J., Shan, S., & Chen, X. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.02957 · Cited by 12
“For AU features, we use the self-supervised 256-dimensional encoder features using TCAE [14], which is trained on the union of VoxCeleb1 [15] and VoxCeleb2 [10] datasets.”
MethodSVEva Fair: A Framework for Evaluating Fairness in Speaker VerificationPreprint
Toussaint, W., & Ding, A. Y. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.12049 · Cited by 17
“Current state-of-the-art approaches use convolutional neural networks, which have been very successful in computer vision tasks, to directly learn speaker embeddings from audio spectrograms [18].”
Qin, X., Li, M., Bu, H., Narayanan, S. S., & Li, H. (2022) · doi:10.21437/ffsvc.2022-3 · Cited by 8
“The fixed training set consists of the following: • VoxCeleb 1&2 [2, 3].”
BackgroundRawSpectrogram: On the Way to Effective Streaming Speech Anti-Spoofing
Grinberg, P., & Shikhov, V. (2023) · IEEE Access · doi:10.1109/access.2023.3321919 · Cited by 8
“A comparison of RECAPA-TDNN with different versions of ECAPA-TDNN is presented in Table 5 in terms of the EER and minimum Detection Cost Function (minDCF, [51]) computed according to the Vox1-O benchmark [52] on the VoxCeleb1 dataset [53].”
Maskrenderer: 3D-infused multi-mask realistic face reenactment
Behrouzi, T., Shahroudnejad, A., & Mousavi, P. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.110891 · Cited by 10
Privacy-preserving speaker verification system using Ranking-of-Element hashing
Nguyen-Le, H.-H., Tran, L., Nguyen, D. S. A., Le‐Khac, N., & Nguyen, T. D. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.111107 · Cited by 8
From One Stolen Utterance: Assessing the Risks of Voice Cloning in the AIGC Era
Wang, K., Chen, M., Lü, L., Feng, J., Chen, Q., Ba, Z., Ren, K., & Chen, C. (2025) · IEEE Symposium on Security and Privacy (S&P) · doi:10.1109/sp61157.2025.00238 · Cited by 27
“Both systems are trained on Voxceleb 1/2 [59], [21] datasets using AM-Softmax Loss [60].”
Disentangling Speech Representations Learning With Latent Diffusion for Speaker Verification
Li, Z., Mak, M.-W., Chien, J.-T., Pilancı, M., Jin, Z., & Meng, H. M. L. (2025) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/taslpro.2025.3610023 · Cited by 11
Post-training for Deepfake Speech Detection
Ge, W., Wang, X., Liu, X., & Yamagishi, J. (2025) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru65441.2025.11434709 · Cited by 36
“…most deepfake datasets are dominated by English and Mandarin Chinese, we also included recent multilingual datasets such as HABLA [32] (Latin American Spanish), Multi-Language Audio Anti-Spoofing Dataset [33] (MLAAD, 38 languages), and SpoofCeleb [34] (multilingual, sourced from VoxCeleb1 [49]) 3 .”
A Robust Approach for Speaker Identification Using Dialect Information
Shah, S. M., Moinuddin, M., & Khan, R. A. (2022) · Applied Computational Intelligence and Soft Computing · doi:10.1155/2022/4980920 · Cited by 8
Towards End-to-End Private Automatic Speaker Recognition
Teixeira, F., Abad, A., Raj, B., & Trancoso, I. M. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-10672 · Cited by 8
“It is composed of short clips taken from interviews uploaded to YouTube [40, 41].”
BackgroundMAST: Multiscale Audio Spectrogram Transformers
Ghosh, S., Seth, A., Umesh, S., & Manocha, D. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095023 · Cited by 16
Model-Agnostic Meta-Learning for Fast Text-Dependent Speaker Embedding Adaptation
Lin, W., & Mak, M. (2023) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2023.3275029 · Cited by 9
“development set and the VoxCeleb2 development set [38], [39].”
BackgroundAutomatic speaker diarization for natural conversation analysis in autism clinical trials
O’Sullivan, J., Bogaarts, G., Schoenenberger, P., Tillmann, J., Slater, D., Mesgarani, N., Eule, E., Kilchenmann, T., Murtagh, L., Hipp, J. F., et al. (2023) · Scientific Reports · doi:10.1038/s41598-023-36701-4 · Cited by 10
“With the advent of deep learning, speaker diarization has become more successful in recent years 14, 15 , in part due to extremely large and publicly available training data (e.g. VoxCeleb 16 ).”
BackgroundPark, S., Mark, M., Park, B., & Hong, H. (2023) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2023.041332 · Cited by 12
3D-aware Blending with Generative NeRFs
Kim, H., Lee, G., Choi, Y., Kim, J.-H., & Zhu, J.-Y. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.02094 · Cited by 14
“However, these methods are limited to faces or hair, and they often need 3D face morphable models [5, 30], or multi-view images [55, 45] to provide 3D information.”
BackgroundWASD: A Wilder Active Speaker Detection Dataset
Roxo, T., Costa, J. C., Inácio, P. R. M., & Proença, H. P. M. C. (2024) · IEEE Transactions on Biometrics Behavior and Identity Science · doi:10.1109/tbiom.2024.3412821 · Cited by 14
One Model to Rule Them all: A Universal Transformer for Biometric Matching
Abdrakhmanova, M., Yermekova, A., Barko, Y., Ryspayev, V., Jumadildayev, M., & Varol, H. A. (2024) · IEEE Access · doi:10.1109/access.2024.3426602 · Cited by 10
“…to the multimodal case as : The multimodal prototypical network network loss (MNPL) aggregates all possible combinations of modality settings : The model was trained and evaluated on two publicly available multimodal datasets suitable for biometric matching: SpeakingFaces [48] and VoxCeleb [49].”
BackgroundAudio-Visual Person Verification Based on Recursive Fusion of Joint Cross-Attention
Praveen, R. G., & Alam, J. (2024) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg59268.2024.10582018 · Cited by 13
“A) Dataset and Evaluation Metrics: We have evaluated the proposed model on the Voxceleb1 dataset [27], obtained from YouTube videos under challenging environments.”
MethodPerceptual Evaluation of Audio-Visual Synchrony Grounded in Viewers’ Opinion Scores
Goncalves, L., Mathur, P., Lavania, C., Cekic, M., Federico, M., & Han, K. J. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72986-7_17 · Cited by 14
Online Neural Speaker Diarization With Target Speaker Tracking
Wang, W., & Li, M. (2024) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2024.3507559 · Cited by 12
A lightweight speaker verification approach for autonomous vehicles
Salah, Y., Shalash, O., & Khatab, E. (2024) · Robotics : Integration, Manufacturing and Control · doi:10.21622/rimc.2024.01.2.1112 · Cited by 8
Privacy Versus Emotion Preservation Trade-Offs in Emotion-Preserving Speaker Anonymization
Cai, Z., Xinyuan, H. L., Garg, A., Garcia, L. P., Duh, K., Khudanpur, S. P., Andrews, N., & Wiesner, M. (2024) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt61566.2024.10832351 · Cited by 20
“We also utilize the Vox-Celeb1 [23] dataset in our study, with the training data including 148,642 utterances from 1,211 speakers and the test set comprising 4,874 utterances from 40 speakers.”
Computer vision-aided audio dataset generation for recognizing construction equipment actions
Jeong, G., Park, M., & Ahn, C. R. (2025) · Automation in Construction · doi:10.1016/j.autcon.2025.106014 · Cited by 10
On Deep Speaker Embeddings for Speaker Verification
Jakubec, M., Jarina, R., Lieskovská, E., & Chmulík, M. (2021) · International Conference on Telecommunications and Signal Processing (TSP) · doi:10.1109/tsp52935.2021.9522589 · Cited by 8
Bidirectional Multiscale Feature Aggregation for Speaker Verification
Qi, J., Guo, W., & Gu, B. (2021) · Interspeech (USB) · doi:10.21437/interspeech.2021-111 · Cited by 8
Borgstrom, B. J. (2021) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2021.3130980 · Cited by 9
SI-Net: Multi-Scale Context-Aware Convolutional Block for Speaker Verification
Li, Z., Fang, C., Xiao, R., Wang, W., & Yan, Y. (2021) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru51503.2021.9688119 · Cited by 8
Robust Automatic Speaker Identification System Using Shuffled MFCC Features
Barhoush, M., Hallawa, A., & Schmeink, A. (2021) · IEEE International Conference on Machine Learning and Applied Network Technologies (ICMLANT) · doi:10.1109/icmlant53170.2021.9690530 · Cited by 8
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.