PP423 - Perceptual Effects of Formant Enhancement with the Factors of Phonetic Type, Listening Conditions, and Language Experience of Listeners

Abstract:
This study examined the effectiveness of second formant (F2) enhancement in improving speech identification across phonetic types (vowels, consonants), listening conditions (speech-shaped noise, six-talker babble at −10 dB and −15 dB SNRs), and listener language experience (native and nonnative English speakers). Participants identified English phonemes with and without F2 enhancement. Results showed significant improvements across various conditions, with greater benefits at −15 dB SNR and for nonnative listeners in vowel identification. These findings demonstrate that F2 enhancement improves speech identification in challenging noise, showing promise as a speech enhancement algorithm.

Summary:
Methods
Native and nonnative English speakers participated in a speech identification task. Participants listened to English vowels and consonants presented with and without F2 enhancement in three listening conditions: quiet, long-term speech-shaped noise (LTSSN), and six-talker babble (6-TB) at SNRs of −10 dB and −15 dB. Stimuli were presented in a randomized order, and participants identified each token using a multiple-choice response format.

Results & Conclusions
F2 enhancement significantly improved identification accuracy for both vowels and consonants across various noise conditions and listener groups. Greater benefits were observed at the more challenging SNR of −15 dB compared to −10 dB, indicating the efficacy of F2 enhancement in severe noise. Nonnative listeners exhibited larger improvements in vowel identification than native listeners, suggesting that F2 enhancement may address specific perceptual challenges faced by nonnative speakers. The improvement was comparable across LTSSN and 6-TB conditions, highlighting the robustness of F2 enhancement across noise types.

These findings demonstrate that F2 enhancement is a promising speech enhancement strategy for improving phonetic identification in adverse noise environments. Its potential to benefit nonnative listeners, who represent a significant population in multilingual societies, underscores its relevance for clinical and technological applications, including hearing devices and speech processing algorithms. Future research could explore the integration of F2 enhancement with other formant-based strategies to optimize speech intelligibility in real-world listening scenarios.

Overall Clarity
The study is well-organized, with clear hypotheses, robust methodology, and results supported by statistical analyses. Conclusions are drawn logically from the findings, emphasizing the clinical and technological implications of the research.

Importance of Work
This work addresses a critical challenge in speech perception and offers an innovative solution with practical implications for improving communication in noisy environments. By focusing on the unique challenges of nonnative listeners, the study has the potential to influence clinical interventions, assistive technology design, and future research on speech enhancement strategies.

Innovation
The study employs a novel application of F2 enhancement across phonetic types, noise conditions, and listener populations, demonstrating its versatility and potential to transform speech processing technologies for diverse users.