Structured Compression and Recovery Fine-Tuning of Egyptian Arabic Hybrid RNN-T/CTC ASR for Offline Raspberry Pi Deployment
Future Internet · 06.11.2026 · 18 (11) · 611
DOI: 10.3390/fi18110611 ↗Kurz gefasst
Fachartikel in der Zeitschrift Future Internet, online erschienen am 06.11.2026. Verfasst von Amir El-Azazy und 4 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Center for Cybersecurity Research and Innovation (CCRI) at SUNY Polytechnic Institute.
Abstract (Originalsprache)
Offline Egyptian Arabic automatic speech recognition (ASR) requires balancing recognition accuracy, model size, and edge device performance. This study investigates structured feed-forward network (FFN) compression and recovery fine-tuning of a 114.6M-parameter FastConformer Hybrid RNN-T/CTC model, followed by offline Raspberry Pi 4 deployment. Reducing FFN width from 2048 to 1792 produced a 105.7M-parameter model, a 7.78% reduction. Structured compression initially degraded recognition accuracy, while subsequent recovery fine-tuning restored performance without restoring the removed FFN channels. On a frozen 10,361-utterance test set, the compressed-and-recovered model achieved 25.01% word error rate (WER) and 11.96% character error rate (CER), compared with 25.88% WER and 12.58% CER for the reference model. A dense control given the same recovery training reached 24.81% WER and 11.87% CER; its paired WER difference from the compressed-and-recovered model was −0.20 percentage points (95% confidence interval −0.44 to +0.04; one training run per model). The earlier advantage over the reference therefore does not indicate a compression-specific benefit. On Raspberry Pi 4, the reference and compressed-and-recovered models achieved aggregate real-time factors of 0.470 and 0.466 and mean latencies of 1.185 and 1.176 s, respectively. The parameter reduction yielded no measured Raspberry Pi speed or memory advantage; traffic-command recognition was not evaluated.
Förderung
- Center for Cybersecurity Research and Innovation (CCRI) at SUNY Polytechnic Institute
Zitieren
APA
El-Azazy, A., Badr, M. M., El-Kammar, R., Abd-Elkader, H., & Fawzy, A. (2026). Structured Compression and Recovery Fine-Tuning of Egyptian Arabic Hybrid RNN-T/CTC ASR for Offline Raspberry Pi Deployment. Future Internet, 18(11), 611. https://doi.org/10.3390/fi18110611
Vancouver
El-Azazy A, Badr MM, El-Kammar R, Abd-Elkader H, Fawzy A. Structured Compression and Recovery Fine-Tuning of Egyptian Arabic Hybrid RNN-T/CTC ASR for Offline Raspberry Pi Deployment. Future Internet. 2026;18(11):611. doi:10.3390/fi18110611
BibTeX
@article{ElAzazy2026,
author = {El-Azazy, Amir and Badr, Mahmoud M. and El-Kammar, Raafat and Abd-Elkader, Hala and Fawzy, Ahmed},
title = {Structured Compression and Recovery Fine-Tuning of Egyptian Arabic Hybrid RNN-T/CTC ASR for Offline Raspberry Pi Deployment},
journal = {Future Internet},
year = {2026},
volume = {18},
number = {11},
pages = {611},
doi = {10.3390/fi18110611}
}
EUVIMED