EUVIMEDEUVIMED entdecken
Crossref

Unified Time–Frequency–Phase Analysis of Speech Using Quaternion Linear Canonical Transform

Mohan Bansal, Kartheek Budime, Varun Gupta

IEEE Signal Processing Letters · 2026 · 33 · 1300-1304

DOI: 10.1109/lsp.2026.3669394 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2026. Verfasst von Mohan Bansal und 2 weiteren Autor:innen. Die Arbeit nennt 22 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The Quaternion Linear Canonical Transform (QLCT) provides a unified time–frequency–phase representation for multidimensional signals.However, its applications have largely been limited to theoretical studies and image processing, leaving speech an inherently one-dimensional non-stationary signal unexplored. This paper presents a Discrete QLCT (DQLCT) implementation that operates directly on real-valued temporal signals by mapping them into the quaternion domain via analytic signal representation. The transform is derived from first principles, with invertibility proven through discrete exponential orthogonality and validated experimentally on a 9.41 s speech signal. Reconstruction error remains at machine precision (mean $8.36 \times 10^{-13}$ ) and Itakura-Saito COSH spectral distances stay below $10^{-12}$ , confirming lossless forward–inverse transform consistency. This establishes DQLCT as a numerically stable, practically implementable tool for quaternion-based speech analysis.

Zitieren

APA

Bansal, M., Budime, K., & Gupta, V. (2026). Unified Time–Frequency–Phase Analysis of Speech Using Quaternion Linear Canonical Transform. IEEE Signal Processing Letters, 33, 1300-1304. https://doi.org/10.1109/lsp.2026.3669394

Vancouver

Bansal M, Budime K, Gupta V. Unified Time–Frequency–Phase Analysis of Speech Using Quaternion Linear Canonical Transform. IEEE Signal Processing Letters. 2026;33:1300-1304. doi:10.1109/lsp.2026.3669394

BibTeX
@article{Bansal2026,
  author = {Bansal, Mohan and Budime, Kartheek and Gupta, Varun},
  title = {Unified Time–Frequency–Phase Analysis of Speech Using Quaternion Linear Canonical Transform},
  journal = {IEEE Signal Processing Letters},
  year = {2026},
  volume = {33},
  pages = {1300--1304},
  doi = {10.1109/lsp.2026.3669394}
}
Kostenlos starten →