EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Spatial-Temporal Relation Enhancement for Speech Emotion Recognition from Acoustic Signals Using Fibonacci Encoding with Diverse Feature Fusion

Shicong Huang, Jinghao Zhang iD, Jingchao Xu, Qi Zhang iD, Zihan Li, Songyang Wang, Zirui Qiu, Zijia Xiong, Jintian Liang, Changzeng Fu

Sensors · 09.08.2026 · 26 (16) · 5050

DOI: 10.3390/s26165050 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 09.08.2026. Verfasst von Shicong Huang und 9 weiteren Autor:innen. Beteiligte Einrichtung: Sydney Smart Technology College, Northeastern University at Qinhuangdao, Qinhuangdao 066004, China. Hauptthemen laut MEDLINE: Emotions, Speech, Acoustics. Die Arbeit nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National College Students' Innovation and Entrepreneurship Training Program, College Students' Innovation and Entrepreneurship Training Program of Northeastern University at Qinhuangdao, National College Students’ Innovation and Entrepreneurship Training Program.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Speech emotion recognition (SER) infers affective states from speech signals, but positional encoding for acoustic tokens remains underexplored in Transformer-based SER. Existing models often reuse encodings designed for text and do not explicitly account for the different sequential and two-dimensional structures of acoustic representations. We propose Fibonacci Position Embedding (FPE) and Fibonacci Target Shutter (FTS). FTS constructs overlapping candidate-index sets over time–frequency token grids, and FPE samples a Fibonacci index and applies its modulo-wrapped, dimension-dependent phase rotation to query and key vectors. The modules are integrated into STRE-Former, which fuses Wav2Vec, log-mel spectrogram, and MFCC representations through asymmetric cross-representation attention with representation-specific positional encodings. We also introduce an implementation-consistent conditional-entropy formulation that quantifies uncertainty in recovering a token location from its sampled positional representation; this quantity characterizes positional ambiguity rather than downstream modeling capacity. Experiments over 64 positional-encoding combinations on IEMOCAP and MELD identify dataset-dependent highest-observed configurations, reaching 74.21% weighted accuracy on IEMOCAP-4, 74.54% on IEMOCAP-6, and 49.44% on MELD. These empirical observations suggest that the relative behavior of positional-encoding strategies may depend on the acoustic representation and evaluation dataset, rather than supporting a single universally optimal scheme.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

EmotionsSpeechAcousticsHumansAlgorithmsSignal Processing, Computer-Assisted

Stichwörter der Autor:innen

Fibonacci encodingTransformeracoustic sensingacoustic signal processingaffective computingmulti-representation fusionpositional embeddingspeech emotion recognition

Autor:innen und Institutionen

  • Sydney Smart Technology College, Northeastern University at Qinhuangdao, Qinhuangdao 066004, China.

ORCID-Profile: Jinghao Zhang, Qi Zhang

Förderung

  • National College Students' Innovation and Entrepreneurship Training Program
  • College Students' Innovation and Entrepreneurship Training Program of Northeastern University at Qinhuangdao
  • National College Students’ Innovation and Entrepreneurship Training Program
  • College Students’ Innovation and Entrepreneurship Training Program of Northeastern University at Qinhuangdao

Zitieren

APA

Huang, S., Zhang, J., Xu, J., Zhang, Q., Li, Z., Wang, S., Qiu, Z., Xiong, Z., Liang, J., & Fu, C. (2026). Spatial-Temporal Relation Enhancement for Speech Emotion Recognition from Acoustic Signals Using Fibonacci Encoding with Diverse Feature Fusion. Sensors, 26(16), 5050. https://doi.org/10.3390/s26165050

Vancouver

Huang S, Zhang J, Xu J, Zhang Q, Li Z, Wang S, et al. Spatial-Temporal Relation Enhancement for Speech Emotion Recognition from Acoustic Signals Using Fibonacci Encoding with Diverse Feature Fusion. Sensors. 2026;26(16):5050. doi:10.3390/s26165050

BibTeX
@article{Huang2026,
  author = {Huang, Shicong and Zhang, Jinghao and Xu, Jingchao and Zhang, Qi and Li, Zihan and Wang, Songyang and Qiu, Zirui and Xiong, Zijia and Liang, Jintian and Fu, Changzeng},
  title = {Spatial-Temporal Relation Enhancement for Speech Emotion Recognition from Acoustic Signals Using Fibonacci Encoding with Diverse Feature Fusion},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {16},
  pages = {5050},
  doi = {10.3390/s26165050}
}
Kostenlos starten →