The DOI matches its Crossref record; 37 of the 37 references it cites were found in a registry or an index.
Extracting Efficient Spectrograms From MP3 Compressed Speech Signals for Synthetic Speech Detection
Open accessAbstract
Many speech signals are compressed with MP3 to reduce the data rate. In many synthetic speech detection methods the spectrogram of the speech signal is used. This usually requires the speech signal to be fully decompressed. We show that the design of MP3 compression allows one to approximate the spectrogram of the MP3 compressed speech efficiently without fully decoding the compressed speech. We denote the spectograms obtained using our proposed approach by Efficient Spectrograms (E-Specs). E-Spec can reduce the complexity of spectrogram computation by ~77.60 percentage points (p.p.) and save ~37.87 p.p. of MP3 decoding time. E-Spec bypasses the reconstruction artifacts introduced by the MP3 synthesis filterbank, which makes it useful in speech forensics tasks. We tested E-Spec in the synthetic speech detection, where a detector is asked to determine whether a speech signal is synthesized or recorded from a human. We examined 4 different neural network architectures to evaluate the performance of E-Spec compared to speech features extracted from the fully decoded speech signal. E-Spec achieved the best synthetic speech detection performance for 3 architectures; it also achieved the best overall detection performance across architectures. The computation of E-Spec is an approximation to Short Time Fourier Transform (STFT). E-Spec can be extended to other audio compression methods.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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 source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Experimental results demonstrate that when using the same backend model, E-Specs outperform other handcrafted features.”
From Audio Deepfake Detection: What Has Been Achieved and What Lies Ahead (2025) · cites this paper as [173]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“This approach enables the direct extraction of spectral features from the MP3 compression domain without full decoding, leading to reduced decoding time and significantly lower computational complexity.”
From Audio Deepfake Detection: What Has Been Achieved and What Lies Ahead (2025) · cites this paper as [173]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
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 · cited 2×
“We changed the values of ′ , , and in the computation of E-Spec, and used the E-Spec to train synthetic speech detectors based on ResNet50 [10].”
MethodAdam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 1×
“During fine-tuning, the models were trained on the training set using the Adam optimizer [17] with a learning rate of 10−5.”
MethodImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380
An introduction to ROC analysis
Fawcett, T. (2006) · Pattern Recognition Letters · doi:10.1016/j.patrec.2005.10.010 · Cited by 21,879 · cited 1×
“The first metric is Area Under Curve (AUC) score, which is defined as the area under the Receiver Operating Characteristics (ROC) curve [7].”
International Organization for Standardization
(2011) · SpringerReference · doi:10.1007/springerreference_16847 · Cited by 12,534
Aggregated Residual Transformations for Deep Neural Networks
Xie, S., Girshick, R., Dollar, P., Tu, Z., & He, K. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.634 · Cited by 12,465 · cited 2×
“For synthetic speech detection, we selected four popular neural network architectures to classify the extracted speech features: ResNet50 [10], EfficientNetV2 [33], ResNeXt50 [38], andMobileNetV3 [11].”
MethodHoward, A., Sandler, M., Chen, B., Wang, W., Chen, L.-C., Tan, M., Chu, G., Vasudevan, V., Zhu, Y., Pang, R., et al. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00140 · Cited by 9,708 · cited 1×
“For synthetic speech detection, we selected four popular neural network architectures to classify the extracted speech features: ResNet50 [10], EfficientNetV2 [33], ResNeXt50 [38], andMobileNetV3 [11].”
MethodProceedings IEEE International Conference on Acoustics Speech and Signal Processing
Zheng, F., McLaughlin, S., & Mulgrew, B. (1991) · Cited by 1,905
Calculation of a constant Q spectral transform
Brown, J. C. (1991) · The Journal of the Acoustical Society of America · doi:10.1121/1.400476 · Cited by 884 · cited 2×
“The computation of E-Spec can be formulated as a fast approximation to the Short Time Fourier Transform (STFT), which is the foundation for many other signal features such as Mel-Frequency Cepstrum Coefficients (MFCC) [43] and Constant-Q Transform (CQT) [6].”
MethodPerceptual coding of digital audio
Painter, T., & Spanias, A. (2000) · Proceedings of the IEEE · doi:10.1109/5.842996 · Cited by 759 · cited 2×
“Because the analysis filterbank is used in many audio compression methods [23], this approach can be extended to other audio compression techniques.”
Method
Research works printed but not linked: 5
- Joebert S. Jacaba. 2001. Audio Compression Using Modified Discrete Cosine Transform: The MP3 Coding Standard. Bachelor's thesis. University of the Philippines, Manila. https://www.math.utah.edu/ gustafso/s2016/2270/project-ideas/audio-mp3-compression-MDCT-jacaba_main.pdf
- 10.1016/B0-12-227410-5 · doi:10.1016/b0-12-227410-5
- Praveen Sripada. 2006. MP3 Decoder in Theory and Practice. Master's thesis. Blekinge Institute of Technology, Ronneby, Sweden. https://www.diva-portal.org/smash/get/diva2:830195/FULLTEXT01.pdf
- Mingxing Tan and Quoc Le. 2021. EfficientNetV2: Smaller Models and Faster Training. Proceedings of International Conference on Machine Learning, Vol. 139 (2021), 10096--10106. Virtual.
- 10.48550/arXiv.2109.00537 · DOI
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
Audio 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
“Xiang et al. [173,174] further proposed Efficient Spectrograms (E-Specs), a novel feature extraction method specifically designed for MP3-compressed speech.”
FairSSD: Understanding Bias in Synthetic Speech Detectors
Singh Yadav, A. K., Bhagtani, K., Salvi, D., Bestagini, P., & Delp, E. J. (2024) · 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw63382.2024.00445 · Cited by 12
“To prevent the malicious use of synthetic speech, existing works have proposed several synthetic speech detec-Figure tors [14–20].”
BackgroundSelf-citationPS3DT: Synthetic Speech Detection Using Patched Spectrogram Transformer
Yadav, A. K. S., Xiang, Z., Bhagtani, K., Bestagini, P., Tubaro, S., & Delp, E. J. (2023) · International Conference on Machine Learning and Applications · doi:10.1109/icmla58977.2023.00075 · Cited by 14
“The horizontal axis represents time and the vertical axis represents frequency [24], [25].”
BackgroundSelf-citationDSVAE: Disentangled Representation Learning for Synthetic Speech Detection
Yadav, A. K. S., Bhagtani, K., Xiang, Z., Bestagini, P., Tubaro, S., & Delp, E. J. (2023) · International Conference on Machine Learning and Applications · doi:10.1109/icmla58977.2023.00072 · Cited by 6
“The Fourier Transform can be used to convert a time domain speech signal into an image representation known as a spectrogram [28], which has been used for speech forensics using a transformer neural network or a Convolutional Neural Network (CNN) [12]–[15], [29], [30].”
MethodSelf-citation
Metadata
- Authors
- Xiang, Z., Yadav, A. K. S., Tubaro, S., Bestagini, P., & Delp, E. J.
- Affiliations
- Purdue University West Lafayette (US)
- Politecnico di Milano (IT)
- Venue
- Proceedings of the 2023 ACM Workshop on Information Hiding and Multimedia Security
- Publisher
- ACM
- Published
- 2023-06-28
- Pages
- 163-168
- OpenAlex
- W4381161978
- Cited by
- 6
- References
- 36
- Citations by year
- 2025: 2 · 2024: 1 · 2023: 3
- FWCI
- 1.07
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- ACM Workshop on Information Hiding and Multimedia Security (IH&MMSec) (publishedVersion, cc-by)
- Virtual Community of Pathological Anatomy (University of Castilla La Mancha) (submittedVersion)
- Funders
- DARPA/AFRL (FA8750-20-2-1004)
- 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
- MP3 compression, synthetic speech detection, speech forensics, audio compression, short-time Fourier transform, computational complexity reduction
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Extracting Efficient Spectrograms From MP3 Compressed Speech Signals for Synthetic Speech Detection. https://app.cochat.ai/papers/10.1145/3577163.3595104
- The paper
- Xiang, Z., Yadav, A. K. S., Tubaro, S., Bestagini, P., & Delp, E. J. (2023). Extracting Efficient Spectrograms From MP3 Compressed Speech Signals for Synthetic Speech Detection. Proceedings of the 2023 ACM Workshop on Information Hiding and Multimedia Security. https://doi.org/10.1145/3577163.3595104
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.