Spoofed Voice Detection using Dense Features of STFT and MDCT Spectrograms

Saleem, Summra; Dilawari, Aniqa; Khan, Usman

doi:10.1109/icai52203.2021.9445259

Cited by 7 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another approach [13] utilizes the Short-Time-Fourier transform (STFT) and a new convolutional neural network for lung cancer diagnosis from spectrochemical analysis. Similarly but in a different domain, in [16] authors utilize a deep neural network to detect audio operations using two types of transform techniques: STFT and Modified Discrete Transform. They improve audio authentication for forensics, successfully recognizing spoofed voices with high accuracy.…”

Section: Related Workmentioning

confidence: 99%

Enhancing Air Quality Forecasting Through Deep Learning and Continuous Wavelet Transform

Manganelli Conforti,

Fanti,

Nardelli

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Air quality forecasting plays a crucial role in environmental management and public health. In this paper, we propose a novel approach that combines deep learning techniques with the Continuous Wavelet Transform (CWT) for air quality forecasting based on sensor data. The proposed methodology is agnostic to the target pollutant and can be applied to estimate any available pollutant without loss of generality. The pipeline consists of two main steps: the generation of stacked samples from raw sensor signals using CWT, and the prediction through a custom deep neural network based on the ResNet18 architecture. We compare our approach with traditional one-dimensional signal processing models. The results show that our 2D pipeline, employing the Morlet mother wavelet, outperforms the baselines significantly. The localized time-frequency representations obtained through CWT highlight hidden dynamics and relationships within the parameter behavior and external factors, leading to more accurate predictions. Overall, our approach demonstrates the potential to advance air quality forecasting and environmental management for healthier living environments worldwide.

show abstract

Section: Related Workmentioning

confidence: 99%

Enhancing Air Quality Forecasting Through Deep Learning and Continuous Wavelet Transform

Manganelli Conforti,

Fanti,

Nardelli

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Although there are many audio tamper detection methods, the technology using electronic network frequency ENF is widely used in multimedia forensics (Zakariah et al, 2018). Although ENF is ideally a sinusoidal signal that oscillates at a nominal frequency, the actual ENF signal will fluctuate slightly with the change of energy supply and the load of the power grid (Saleem et al, 2021). When a segment is inserted or deleted from a recording, the ENF of that segment also changes.…”

Section: Research On Tamper Detection Based On Electronic Network Fre...mentioning

confidence: 99%

“…Chen (Chen et al, 2016) detects tampered audio in the time domain through discrete wavelet packet decomposition and singularity analysis of speech signals for audio tampering operations of insertion and deletion. Saleem (Saleem et al, 2021) inputted the Short-Time Fourier transform (STFT) and Modified Discrete Cosine Transform (MDCT) spectra of audio into the convolutional neural network to identify Spoofed Voices. There is also much research on copy-move Forgery of the copy-paste type.…”

Section: Research On Tamper Detection Based On Audio Speech Contentmentioning

confidence: 99%

Audio Tampering Forensics Based on Representation Learning of ENF Phase Sequence

Zeng

Yang

Wang

et al. 2022

International Journal of Digital Crime and Forensics

View full text Add to dashboard Cite

This paper proposes an audio tampering detection method based on the ENF phase and BI-LSTM network from the perspective of temporal feature representation learning. First, the ENF phase is obtained by discrete Fourier transform of ENF component in audio. Second, the ENF phase is divided into frames to obtain ENF phase sequence characterization, and each frame is represented as the change information of the ENF phase in a period. Then, the BI-LSTM neural network is used to train and output the state of each time step, and the difference information between real audio and tampered audio is obtained. Finally, these differences were fitted and dimensionally reduced by the fully connected network and classified by the Softmax classifier. Experimental results show that the performance of this method is better than the state-of-the-art approaches.

show abstract

“…The undetectable manipulation of digital speech avatars poses substantial threats to judicial processes, political fields, and social security. Contemporary speech forensics techniques are pivotal in ensuring the integrity of digital avatars and focus on detecting tampering facilitated by audio editing software, such as deletion, insertion, copy and move, splicing, resampling and recompression of audio clips [4][5][6][7]. It is worth noting that in the field of speech content forensics, there are more forensic methods for speech deletion, copy and move, splicing, and other tampering approaches [8][9][10], while there are relatively few methods for speech resampling forensics, and these tampering means are often accompanied by resampling operations.…”

Section: Introductionmentioning

confidence: 99%

Pyramid Feature Attention Network for Speech Resampling Detection

Zhou,

Zhang,

Wang

et al. 2024

Applied Sciences

View full text Add to dashboard Cite

Speech forgery and tampering, increasingly facilitated by advanced audio editing software, pose significant threats to the integrity and privacy of digital speech avatars. Speech resampling is a post-processing operation of various speech-tampering means, and the forensic detection of speech resampling is of great significance. For speech resampling detection, most of the previous works used traditional methods of feature extraction and classification to distinguish original speech from forged speech. In view of the powerful ability of deep learning to extract features, this paper converts the speech signal into a spectrogram with time-frequency characteristics, and uses the feature pyramid network (FPN) with the Squeeze and Excitation (SE) attention mechanism to learn speech resampling features. The proposed method combines the low-level location information and the high-level semantic information, which dramatically improves the detection performance of speech resampling. Experiments were carried out on a resampling corpus made on the basis of the TIMIT dataset. The results indicate that the proposed method significantly improved the detection accuracy of various resampled speech. For the tampered speech with a resampling factor of 0.9, the detection accuracy is increased by nearly 20%. In addition, the robustness test demonstrates that the proposed model has strong resistance to MP3 compression, and the overall performance is better than the existing methods.

show abstract

Spoofed Voice Detection using Dense Features of STFT and MDCT Spectrograms

Cited by 7 publications

References 16 publications

Enhancing Air Quality Forecasting Through Deep Learning and Continuous Wavelet Transform

Enhancing Air Quality Forecasting Through Deep Learning and Continuous Wavelet Transform

Audio Tampering Forensics Based on Representation Learning of ENF Phase Sequence

Pyramid Feature Attention Network for Speech Resampling Detection

Contact Info

Product

Resources

About