A Comparative Study on Physical and Perceptual Features for Deepfake Audio Detection
The DOI matches its Crossref record; 27 of the 27 references it cites were found in a registry or an index.
Abstract
Audio content synthesis has stepped into a new era and brought a great threat to daily life since the development of deep learning techniques. The ASVSpoof Challenge and the ADD Challenge have been launched to motivate the development of Deepfake audio detection algorithms. Currently, the detection models, which consist of front-end feature extractors and back-end classifiers, utilize the physical features mainly, rather than the perceptual features that relate to natural emotions or breathiness. Therefore, we provide a comprehensive study on 16 physical and perceptual features and evaluate their effectiveness in both Track 1 and Track 2 of the ADD Challenge. Based on results, PLP, as a perceptual feature, outperforms the rest of the features in Track 1, while CQCC has the best performance in Track 2. Our experiments demonstrate the significance of perceptual features in detecting Deepfake audios. We also seek to explore the underlying characteristics of features that can distinguish Deepfake audio from a real one. We perform statistical analysis on each feature to show its distribution differences on real and synthesized audios. This paper will provide a potential direction in selecting appropriate feature extraction methods for the future implementation of detection models.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“Handcrafted-feature-based methods utilize expert-based audio processing techniques as front-end to convert the audio waveform into predefined acoustic features.”
From Deepfake detection across image, video, and audio: a comprehensive survey with empirical evaluation of generalization and robustness (2026) · cites this paper as [130]
In the citing paper
“a significant body of research [64,96,97,98,99] has demonstrated that the performance of detection models largely depends on the choice of input features.”
From Audio Deepfake Detection: What Has Been Achieved and What Lies Ahead (2025) · cites this paper as [99]
Citations
Cites
Davis, S., & Mermelstein, P. (1980) · IEEE Transactions on Acoustics, Speech, and Signal Processing · doi:10.1109/tassp.1980.1163420 · Cited by 5,465
librosa: Audio and Music Signal Analysis in Python
McFee, B., Raffel, C., Liang, D., Ellis, D., McVicar, M., Battenberg, E., & Nieto, O. (2015) · Proceedings of the Python in Science Conference · doi:10.25080/majora-7b98e3ed-003 · Cited by 3,033
Speaker Recognition from Raw Waveform with SincNet
Ravanelli, M., & Bengio, Y. (2018) · 2018 IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt.2018.8639585 · Cited by 896
Introducing Parselmouth: A Python interface to Praat
Jadoul, Y., Thompson, B., & de Boer, B. (2018) · Journal of Phonetics · doi:10.1016/j.wocn.2018.07.001 · Cited by 506
Constant Q Cepstral Coefficients: A Spoofing Countermeasure for Automatic Speaker Verification
Todisco, M., Delgado, H., & Evans, N. (2017) · Computer Speech & Language · doi:10.1016/j.csl.2017.01.001 · Cited by 424
Vocal Acoustic Analysis – Jitter, Shimmer and HNR Parameters
Teixeira, J. P., Oliveira, C., & Lopes, C. (2013) · Procedia Technology · doi:10.1016/j.protcy.2013.12.124 · Cited by 412
End-to-End anti-spoofing with RawNet2
Tak, H., Patino, J., Todisco, M., Nautsch, A., Evans, N., & Larcher, A. (2021) · ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp39728.2021.9414234 · Cited by 402 · cited 2×
“RawNet2[27]: The RawNet2 is a deep learning-based Deepfake audio detection algorithm, which consists of CNN and ResNet structures.”
MethodA comparison of features for synthetic speech detection
Sahidullah, M., Kinnunen, T., & Hanilçi, C. (2015) · Interspeech 2015 · doi:10.21437/interspeech.2015-472 · Cited by 382 · cited 2×
“[23] compare 19 acoustic features based on first-order Fourier coefficients and second-order spectral features.”
BackgroundFerrand, C. T. (2002) · Journal of Voice · doi:10.1016/s0892-1997(02)00123-6 · Cited by 259
Detecting Deep-Fake Videos from Appearance and Behavior
Agarwal, S., Farid, H., El-Gaaly, T., & Lim, S.-N. (2020) · 2020 IEEE International Workshop on Information Forensics and Security (WIFS) · doi:10.1109/wifs49906.2020.9360904 · Cited by 242 · cited 1×
“As audio synthesis techniques grow in popularity, especially after the development of deep learning techniques, anti-spoofing methods are proposed in the literature to detect bonafide audio from Deepfake ones [1, 2, 31, 36].”
MethodRecurrent convolutional structures for audio spoof and video deepfake detection
Chintha, A., Thai, B., Sohrawardi, S. J., Bhatt, K., Hickerson, A., Wright, M., & Ptucha, R. (2020) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2020.2999185 · Cited by 226 · cited 1×
“In End-to-End architectures, the network accepts raw audio as input and the utterances are processed through the network to produce the classification outcome [7, 27].”
MethodReplay and Synthetic Speech Detection with Res2Net Architecture
Li, X., Li, N., Weng, C., Liu, X., Su, D., Yu, D., & Meng, H. (2021) · ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp39728.2021.9413828 · Cited by 186 · cited 1×
“SE-Res2Net50[18]: This classifier is an upgraded ResNet detection architecture, which enables multiple feature scales during the training process.”
MethodGeneralization of Spectral Flatness Measure for Non-Gaussian Linear Processes
Dubnov, S. (2004) · IEEE Signal Processing Letters · doi:10.1109/lsp.2004.831663 · Cited by 179
ON THE USE OF ZERO-CROSSING RATE FOR AN APPLICATION OF CLASSIFICATION OF PERCUSSIVE SOUNDS
Gouyon, F., Pachet, F., & Delerue, O. (2000) · Cited by 175
ADD 2022: the first Audio Deep Synthesis Detection Challenge
Yi, J., Fu, R., Tao, J., Nie, S., Ma, H., Wang, C., Wang, T., Tian, Z., Bai, Y., Fan, C., et al. (2022) · ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp43922.2022.9746939 · Cited by 165
ASVspoof 2021: Automatic Speaker Verification Spoofing and Countermeasures Challenge Evaluation Plan
Delgado, H., Evans, N., Kinnunen, T., Lee, K. A., Liu, X., Nautsch, A., Patino, J., Sahidullah, M., Todisco, M., Wang, X., et al. (2021) · arXiv · doi:10.48550/arxiv.2109.00535 · Cited by 129 · cited 1×
“Both LA and DF sets consist of synthesized audio samples generated by Text-to-Speech (TTS) and Voice Conversion (VC) algorithm, while the DF task involves the compressed audios without ASV systems [11].”
MethodSome Commonly Used Speech Feature Extraction Algorithms
Ajibola Alim, S., & Khair Alang Rashid, N. (2018) · From Natural to Artificial Intelligence - Algorithms and Applications · doi:10.5772/intechopen.80419 · Cited by 119
Auditory Scene Analysis: An Attention Perspective
Sussman, E. S. (2017) · Journal of Speech Language and Hearing Research · doi:10.1044/2017_jslhr-h-17-0041 · Cited by 100
Toward Robust Audio Spoofing Detection: A Detailed Comparison of Traditional and Learned Features
Balamurali, B. T., Lin, K. E., Lui, S., Chen, J.-M., & Herremans, D. (2019) · IEEE Access · doi:10.1109/access.2019.2923806 · Cited by 75
Long Range Acoustic and Deep Features Perspective on ASVspoof 2019
Das, R. K., Yang, J., & Li, H. (2019) · 2019 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU) · doi:10.1109/asru46091.2019.9003845 · Cited by 65
Extended Constant-Q Cepstral Coefficients for Detection of Spoofing Attacks
Yang, J., Das, R. K., & Li, H. (2018) · 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.23919/apsipa.2018.8659537 · Cited by 53
Das, P. P., Allayear, S. M., Amin, R., & Rahman, Z. (2016) · International Conference on Advanced Computational Intelligence (ICACI) · doi:10.1109/icaci.2016.7449852 · Cited by 36
Gao, Y., Lian, J., Raj, B., & Singh, R. (2021) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt48900.2021.9383558 · Cited by 31
Pindrop Labs' Submission to the ASVspoof 2021 Challenge
Chen, T., Khoury, E., Phatak, K., & Sivaraman, G. (2021) · 2021 Edition of the Automatic Speaker Verification and Spoofing Countermeasures Challenge · doi:10.21437/asvspoof.2021-14 · Cited by 30 · cited 1×
“Chen et al. [6] apply three variants of ResNet structures to extract deep feature representation of the raw audio signals.”
BackgroundDeep Learning in Face Synthesis: A Survey on Deepfakes
Zhang, T., Deng, L., Zhang, L., & Dang, X. (2020) · 2020 IEEE 3rd International Conference on Computer and Communication Engineering Technology (CCET) · doi:10.1109/ccet50901.2020.9213159 · Cited by 18
An effective, simple tempo estimation method based on self-similarity and regularity
Tzanetakis, G., & Percival, G. (2013) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2013.6637645 · Cited by 11
Voice Conversion and Spoofed Voice Detection from Parallel English and Urdu Corpus using Cyclic GANs
Saleem, S., Dilawari, A., Ghani Khan, M. U., & Husnain, M. (2019) · 2019 International Conference on Robotics and Automation in Industry (ICRAI) · doi:10.1109/icrai47710.2019.8967385 · Cited by 5 · cited 1×
“Perceptual features Pitch based feature Computed with Parselmouth software (2, 105) Onset Strength Hop length = 160, time lag = 1 (1, 301) HNR Computed with Parselmouth software (1, 107) Intensity Computed with Parselmouth software (1, 129) Jitter-shimmer based feature Computed with Parselmouth software (1, 11) [14] C…”
Method
Research works printed but not linked: 8
- Ehab A AlBadawy , Siwei Lyu , and Hany Farid . 2019 . Detecting AI-Synthesized Speech Using Bispectral Analysis .. In CVPR workshops. 104--109 . Ehab A AlBadawy, Siwei Lyu, and Hany Farid. 2019. Detecting AI-Synthesized Speech Using Bispectral Analysis.. In CVPR workshops. 104--109.
- Moustafa Alzantot , Ziqi Wang , and Mani B Srivastava . 2019. Deep residual neural networks for audio spoofing detection. arXiv preprint arXiv:1907.00501 ( 2019 ). Moustafa Alzantot, Ziqi Wang, and Mani B Srivastava. 2019. Deep residual neural networks for audio spoofing detection. arXiv preprint arXiv:1907.00501 (2019). · arXiv
- Hynek Hermansky . 1990. Perceptual linear predictive (PLP) analysis of speech. the Journal of the Acoustical Society of America , Vol. 87 , 4 ( 1990 ), 1738--1752. Hynek Hermansky. 1990. Perceptual linear predictive (PLP) analysis of speech. the Journal of the Acoustical Society of America, Vol. 87, 4 (1990), 1738--1752.
- Zeyan Oo , Longbiao Wang , Khomdet Phapatanaburi , Meng Liu , Seiichi Nakagawa , Masahiro Iwahashi , and Jianwu Dang . 2019. Replay attack detection with auditory filter-based relative phase features. EURASIP journal on audio, speech, and music processing , Vol. 2019 , 1 ( 2019 ), 1--11. Zeyan Oo, Longbiao Wang, Khomdet Phapatanaburi, Meng Liu, Seiichi Nakagawa, Masahiro Iwahashi, and Jianwu Dang. 2019. Replay attack detection with auditory filter-based relative phase features. EURASIP journal on audio, speech, and music processing, Vol. 2019, 1 (2019), 1--11.
- William P Shofner and George Selas . 2002. Pitch strength and Stevens's power law. Perception & psychophysics , Vol. 64 , 3 ( 2002 ), 437--450. William P Shofner and George Selas. 2002. Pitch strength and Stevens's power law. Perception & psychophysics , Vol. 64, 3 (2002), 437--450.
- Ahmet Semih Ucc an, Fatih Mustafa Bucc ak, Mehmet Ali Han Tutuk , Halis. Ibrahim Aydin , Ertuug rul Semiz , and c S erif Bahtiyar . 2021 . Deepfake and Security of Video Conferences. In 2021 6th International Conference on Computer Science and Engineering (UBMK). IEEE, 36--41 . Ahmet Semih Ucc an, Fatih Mustafa Bucc ak, Mehmet Ali Han Tutuk, Halis. Ibrahim Aydin, Ertuug rul Semiz, and cS erif Bahtiyar. 2021. Deepfake and Security of Video Conferences. In 2021 6th International Conference on Computer Science and Engineering (UBMK). IEEE, 36--41.
- Zhenzong Wu , Rohan Kumar Das , Jichen Yang, and Haizhou Li. 2020 . Light convolutional neural network with feature genuinization for detection of synthetic speech attacks. arXiv preprint arXiv:2009.09637 (2020). Zhenzong Wu, Rohan Kumar Das, Jichen Yang, and Haizhou Li. 2020. Light convolutional neural network with feature genuinization for detection of synthetic speech attacks. arXiv preprint arXiv:2009.09637 (2020). · arXiv
- Zhizheng Wu , Tomi Kinnunen , Nicholas Evans , Junichi Yamagishi , Cemal Hanilcc i , Md Sahidullah, and Aleksandr Sizov. 2015 . ASVspoof 2015: the first automatic speaker verification spoofing and countermeasures challenge. In Sixteenth annual conference of the international speech communication association. Zhizheng Wu, Tomi Kinnunen, Nicholas Evans, Junichi Yamagishi, Cemal Hanilcc i, Md Sahidullah, and Aleksandr Sizov. 2015. ASVspoof 2015: the first automatic speaker verification spoofing and countermeasures challenge. In Sixteenth annual conference of the international speech communication association.
Cited by
A Survey on Speech Deepfake Detection
Li, M., Ahmadiadli, Y., & Zhang, X.-P. (2025) · ACM Computing Surveys · doi:10.1145/3714458 · Cited by 101
Self-citationAudio Deepfake Detection: What Has Been Achieved and What Lies Ahead
Zhang, B., Cui, H., Nguyen, V., & Whitty, M. (2025) · Sensors · doi:10.3390/s25071989 · Cited by 72
SafeEar: Content Privacy-Preserving Audio Deepfake Detection
Li, X., Li, K., Zheng, Y., Yan, C., Ji, X., & Xu, W. (2024) · Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3658644.3670285 · Cited by 39
From Audio Deepfake Detection to AI-Generated Music Detection–A Pathway and Overview
Li, Y., Milling, M., Specia, L., & Schuller, B. W. (2026) · IEEE Computational Intelligence Magazine · doi:10.1109/mci.2026.3721724 · Cited by 36
Firc, A., Malinka, K., & Hanáček, P. (2025) · Cybersecurity · doi:10.1186/s42400-024-00346-1 · Cited by 34
Metadata
- Authors
- Li, M., Ahmadiadli, Y., & Zhang, X.-P.
- Affiliations
- Toronto Metropolitan University (CA)
- Venue
- Proceedings of the 1st International Workshop on Deepfake Detection for Audio Multimedia
- Publisher
- ACM
- Published
- 2022-10-10
- Pages
- 35-41
- OpenAlex
- W4298394324
- Cited by
- 35
- References
- 26
- Citations by year
- 2026: 6 · 2025: 16 · 2024: 9 · 2023: 4
- FWCI
- 4.01
- Type
- conference paper
- Language
- en
- Open access
- closed
- 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 › Signal Processing › Music and Audio Processing
- Keywords
- audio deepfake detection, perceptual features, physical characteristics, sound synthesis, PLP, feature extraction, anti-spoofing
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.
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.