PATSE: PyTorch-Native Polar Encoding for Runtime Stability in Audio Processing

Nuk Ghurroh Setyoningrum; Ema Utami; Kusrini Kusrini; Ferry Wahyu Wibowo

doi:10.29207/resti.v10i5.7776

Authors

Nuk Ghurroh Setyoningrum Universitas Amikom Yogyakarta
Ema Utami Universitas Amikom Yogyakarta
Kusrini Kusrini Universitas Amikom Yogyakarta
Ferry Wahyu Wibowo Universitas Amikom Yogyakarta

DOI:

https://doi.org/10.29207/resti.v10i5.7776

Keywords:

audio feature extraction, deep learning, infant cry classification, polarized amplitude time spiral encoding, real-time inference

Abstract

Efficient and stable audio feature extraction is essential for real-time AI systems, yet conventional spectral representations often incur preprocessing overhead and framework-conversion dependencies. This study proposes PATSE (Polarized Amplitude Time Spiral Encoding), a PyTorch-native spatial–temporal encoding framework that converts one-dimensional audio into CNN-compatible two-dimensional polar representations while preserving temporal–amplitude information. Evaluated on 100 infant cry samples, PATSE achieved competitive CPU runtime (5.37 ms), improved latency stability over Mel-Spectrogram, and 91% classification accuracy with a Macro-F1 score of 0.85 using a lightweight CNN. These results suggest that PATSE provides an effective balance between computational efficiency, framework integration, and discriminative representation in the evaluated CPU-based infant cry setting, while broader validation is reserved for future work.

Downloads

Download data is not yet available.

References

[1] N. G. Setyoningrum, E. Utami, Kusrini, and F. W. Wibowo, “A Systematic Literature Review of Audio Signal Processing Methods for Infant Cry Recognition and Interpretation,” CENiM, 2024, doi: 10.1109/CENIM64038.2024.10882830.

[2] T. Ozseven, “Infant cry classification by using different deep neural network models and hand-crafted features,” Biomed. Signal Process. Control, vol. 83, p. 104648, 2023, doi: 10.1016/j.bspc.2023.104648.

[3] R. Jahangir, “CNN‐SCNet: A CNN net‐based deep learning framework for infant cry detection in household setting,” Engineering Reports, 2023, doi: 10.1002/eng2.12786.

[4] N. G. Setyoningrum, E. Utami, K. Kusrini, and F. W. Wibowo, “Improving Infant Cry Recognition Using MFCC And CNN-Based Audio Augmen tation,” Jurnal Teknik Informatika (Jutif), 2025, doi: 10.52436/1.jutif.2025.6.2.4373.

[5] Z. Firas, A. A. Nashaat, and G. Ahmad, “Optimizing Infant Cry Recognition: A Fusion of LPC and MFCC Features in Deep Learning Models,” in International Conference on Advances in Biomedical Engineering, ICABME, Institute of Electrical and Electronics Engineers Inc., 2023, pp. 232–237. doi: 10.1109/ICABME59496.2023.10293083.

[6] H. Choi, L. Zhang, and C. Watkins, “Dual representations: A novel variant of Self-Supervised Audio Spectrogram Transformer with multi-layer feature fusion and pooling combinations for sound classification,” Neurocomputing, vol. 623, Mar. 2025, doi: 10.1016/j.neucom.2025.129415.

[7] A. S. Podda, R. Balia, L. Pompianu, S. Carta, G. Fenu, and R. Saia, “CARgram: CNN-based accident recognition from road sounds through intensity-projected spectrogram analysis,” Digital Signal Processing: A Review Journal, vol. 147, Apr. 2024, doi: 10.1016/j.dsp.2024.104431.

[8] N. G. Setyoningrum, E. Utami, Kusrini, and F. W. Wibowo, “A Comprehensive Survey of Infant Cry Classification Research Trends and Methods: A Systematic Review,” in 2024 6th International Conference on Cybernetics and Intelligent System (ICORIS), IEEE, Nov. 2024, pp. 1–6. doi: 10.1109/ICORIS63540.2024.10903693.

[9] T. Mary Little Flower, T. Jaya, and S. Christopher Ezhil Singh, “Data augmentation using a 1D-CNN model with MFCC/MFMC features for speech emotion recognition,” Automatika, vol. 65, no. 4, pp. 1325–1338, 2024, doi: 10.1080/00051144.2024.2371249.

[10] K. R. Mannem, E. Mengiste, S. Hasan, B. G. de Soto, and R. Sacks, “Smart audio signal classification for tracking of construction tasks,” Autom. Constr., vol. 165, Sep. 2024, doi: 10.1016/j.autcon.2024.105485.

[11] Y. Zayed, A. Hasasneh, and C. Tadj, “Infant Cry Signal Diagnostic System Using Deep Learning and Fused Features,” Diagnostics, vol. 13, no. 12, p. 2107, 2023, doi: 10.3390/diagnostics13122107.

[12] N. Ghurroh Setyoningrum, E. Utami, F. Wahyu Wibowo, “Polarized Amplitude Time Spiral Encoding for Infant Cry Audio Augmentation and Convolutional Neural Network Journal of Innovation Information Technology and Application, vol. 7, no. 2, pp. 379–394, 2025, doi: 10.35970/jinita.v7i2.2960.

[13] M. W. Lakdari, A. H. Ahmad, S. Sethi, G. A. Bohn, and D. J. Clink, “Mel-frequency cepstral coefficients outperform embeddings from pre-trained convolutional neural networks under noisy conditions for discrimination tasks of individual gibbons,” Ecol. Inform., vol. 80, p. 102457, 2024, doi: 10.1016/j.ecoinf.2023.102457.

[14] G. Coro, S. Bardelli, A. Cuttano, R. T. Scaramuzzo, and M. Ciantelli, “A self-training automatic infant-cry detector,” Neural Comput. Appl., vol. 35, no. 11, pp. 8543–8559, Apr. 2023, doi: 10.1007/s00521-022-08129-w.

[15] A. Abbaskhah, H. Sedighi, and H. Marvi, “Infant cry classification by MFCC feature extraction with MLP and CNN structures,” Biomed. Signal Process. Control, vol. 86, Sep. 2023, doi: 10.1016/j.bspc.2023.105261.

[16] Y. Gong, Y. A. Chung, and J. Glass, “Ast: Audio spectrogram transformer,” in Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH, International Speech Communication Association, 2021, pp. 56–60. doi: 10.21437/Interspeech.2021-698.

[17] W.-N. Hsu, B. Bolte, Y.-H. H. Tsai, K. Lakhotia, R. Salakhutdinov, and A. Mohamed, “HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units,” Jun. 2021, [Online]. Available: http://arxiv.org/abs/2106.07447. doi.org/10.1109/TASLP.2021.3122291

[18] K. Zaman, K. Li, M. Sah, C. Direkoglu, S. Okada, and M. Unoki, “Transformers and audio detection tasks: An overview,” Mar. 01, 2025, Elsevier Inc. doi: 10.1016/j.dsp.2024.104956.

[19] X. Li and Z. Zhang, “Cross-feature fusion speech emotion recognition based on attention mask residual network and Wav2vec 2.0,” Digital Communications and Networks, vol. 11, no. 5, pp. 1567–1577, Oct. 2025, doi: 10.1016/j.dcan.2024.10.007.

[20] J. Cai, Y. Song, J. Wu, and X. Chen, “Voice Disorder Classification Using Wav2vec 2.0 Feature Extraction,” Journal of Voice, 2024, doi: 10.1016/j.jvoice.2024.09.002.

[21] W. Chen et al., “Classifying Heart-Sound Signals Based on CNN Trained on MelSpectrum an d Log-MelSpectrum Features,” Bioengineering, vol. 10, no. 6, p. 645, 2023, doi: 10.3390/bioengineering10060645.

[22] T. Guo et al., “SPIRAL: integrating and aligning spatially resolved transcriptomics da ta across different experiments, conditions, and technologies,” Genome Biol., 2023, doi: 10.1186/s13059-023-03078-6.

[23] A. Piemontese, A. Ugolini, M. Morini, G. Colavolpe, and T. Eriksson, “Spiral Constellations for Nonlinear Channels,” IEEE Communications Letters, vol. 28, no. 9, pp. 2016–2020, 2024, doi: 10.1109/lcomm.2024.3438830.

[24] B. Wang, T. Cai, X. Cai, W. Zhou, and Z. Liu, “Polarized imaging dynamic light scattering for simultaneous measurement of nanoparticle size and morphology,” Particuology, vol. 85, pp. 213–223, Feb. 2024, doi: 10.1016/j.partic.2023.06.004.

[25] G. M. Moatimid and T. S. Amer, “Dynamical system of a time-delayed ϕ6-Van der Pol oscillator: a non-perturbative approach,” Sci. Rep., vol. 13, no. 1, 2023, doi: 10.1038/s41598-023-38679-5.

[26] Y. Jing and X. Zhao, “DQ-Former: Querying Transformer with Dynamic Modality Priority for Cognitive-aligned Multimodal Emotion Recognition in Conversation,” in MM 2024 - Proceedings of the 32nd ACM International Conference on Multimedia, Association for Computing Machinery, Inc, Oct. 2024, pp. 4795–4804. doi: 10.1145/3664647.3681599.

[27] A. Laguna et al., “How can cry acoustics associate newborns’ distress levels with neurophysiological and behavioral signals?,” Front. Neurosci., 2023, doi: 10.3389/fnins.2023.1266873.

[28] S. Lahmiri, C. Tadj, C. Gargour, and S. Bekiros, “Optimal tuning of support vector machines and k-NN algorithm by using Bayesian optimization for newborn cry signal diagnosis based on audio signal processing features,” Chaos Solitons Fractals, vol. 167, Feb. 2023, doi: 10.1016/j.chaos.2022.112972.

[29] S. A. Patel and A. Yildirim, “Non-stationary neural signal to image conversion framework for image-b ased deep learning algorithms,” Front. Neuroinform., vol. 17, 2023, doi: 10.3389/fninf.2023.1081160.

[30] N. Chelimilla, V. Chinthapenta, and S. Korla, “Addressing data scarcity using audio signal augmentation and deep learning for bolt looseness prediction,” Smart Mater. Struct., vol. 33, no. 8, p. 085012, 2024, doi: 10.1088/1361-665x/ad5c24.

[31] K. Ashwini, P. M. D. R. Vincent, K. Srinivasan, and C. Y. Chang, “Deep Learning Assisted Neonatal Cry Classification via Support Vector Machine Models,” Front. Public Health, vol. 9, Jun. 2021, doi: 10.3389/fpubh.2021.670352.

[32] M. Zhang, Y. Zheng, X. Maidaiti, B. Liang, Y. Wei, and F. Sun, “Integrating Machine Learning into Statistical Methods in Disease Risk Prediction Modeling: A Systematic Review,” Health Data Science, vol. 4, 2024, doi: 10.34133/hds.0165.

[33] J. Allgaier, L. Mulansky, R. Draelos, and R. Pryss, “How does the model make predictions? A systematic literature review on the explainability power of machine learning in healthcare,” Artif. Intell. Medicine, 2023, doi: 10.1016/j.artmed.2023.102616.

[34] S. Lee, C. Yin, and P. Zhang, “Stable clinical risk prediction against distribution shift in electron ic health records,” Patterns, 2023, doi: 10.1016/j.patter.2023.100828.

[35] A. Ayari, H. Hamdi, and K. A. Alsulbi, “E-health Application in IoMT Environment Deployed in An Edge and Cloud Computing Platforms,” in Procedia Computer Science, Elsevier B.V., 2024, pp. 1019–1028. doi: 10.1016/j.procs.2024.09.521.

[36] T. Hama et al., “Enhancing Patient Outcome Prediction Through Deep Learning with Sequential Diagnosis Codes from Structured Electronic Health Record Data: Sy stematic Review,” J. Med. Internet Res., vol. 27, p. e57358, 2025, doi: 10.2196/57358.