Comprehensive Review Paper on SpeechEase: AI-Based Multilingual Speech Therapy System

This review considered AI-assisted speech technologies relevant to multilingual pronunciation practice. The selected literature demonstrates progress in automatic speech analy-sis, personalized pronunciation correction, multilingual ASR, digital speech practice, and articulatory speech analysis [2], [4]–[6]. It also highlights challenges related to phoneme com-plexity, language coverage, data availability, evidence quality, and system integration [14]–[16]. The review identifies a potential gap in connecting assess-ment, identification of practice needs, personalized exercises, reassessment, and longitudinal progress monitoring in one accessible platform. In response, an initial framework for an AI-assisted multilingual speech-practice application has been outlined. The project remains at the planning stage: no implementation, experiment, or result is claimed in this paper. Future work should refine the requirements, verify the selected literature, assess available datasets and models, and then develop and evaluate a limited prototype. Any eventual system should communicate uncertainty and remain an as-sistive practice tool rather than a substitute for professional clinical assessment.

[1]     “Artificial Intelligence for Speech Classification and Enhancement of Speech and Language Disorders: Techniques, Applications, and Future Directions,” IEEE Access, vol. 13, pp. 177136–177159, 2025, doi: 10.1109/ACCESS.2025.3620114. [2]     M. Carl, E. Rudyk, Y. Shapira, H. Leavy Rusiewicz, and M. Icht, “Accuracy of Speech Sound Analysis: Comparison of an Automatic Artificial Intelligence Algorithm With Clinician Assessment,” Journal of Speech, Language, and Hearing Research, vol. 67, no. 9, pp. 3004–3021, 2024, doi: 10.1044/2024 _JSLHR-24-00009. [3]     N. R. Benway and J. L. Preston, “Artificial Intelligence-Assisted Speech Therapy for /r/: A Single-Case Experimental Study,” American Journal of Speech-Language Pathology, vol. 33, no. 5, pp. 2461–2486, 2024, doi: 10.1044/2024 _AJSLP-23-00448. [4]     R. Lakshminarayanan, A. S. K. Shaik, and A. Balasundaram, “Au-tomated Speech Therapy Through Personalized Pronunciation Cor-rection Using Reinforcement Learning and Large Language Mod-els,” Results in Engineering, vol. 25, Art. no. 103943, 2025, doi: 10.1016/j.rineng.2025.103943. [5]     K. Tripathi, R. Gothi, and P. Wasnik, “Enhancing Whisper’s Accuracy and Speed for Indian Languages through Prompt-Tuning and Tokeniza-tion,” in Proc. IEEE Int. Conf. Acoustics, Speech and Signal Processing (ICASSP), 2025, pp. 1–5, doi: 10.1109/ICASSP49660.2025.10887595. [6]     N. R. Benway, S. Tabatabaee, D. Wang, B. Munson, J. L. Preston, and C. Espy-Wilson, “Perceptual Ratings Predict Speech Inversion Articulatory Kinematics in Childhood Speech Sound Disorders,” Journal of Speech, Language, and Hearing Research, vol. 69, no. 2, pp. 541–561, 2026, doi: 10.1044/2025 _JSLHR-25-00515. [7]     H. Snyder, “Literature review as a research methodology: An overview and guidelines,” Journal of Business Research, vol. 104, pp. 333–339, 2019, doi: 10.1016/j.jbusres.2019.07.039. [8]     T. Ahn *et al.*, “Automatic Speech Recognition (ASR) for the Diagnosis of Pronunciation of Speech Sound Disorders in Korean Children,” Clinical Linguistics & Phonetics, vol. 39, no. 10, pp. 913–926, 2025, doi: 10.1080/02699206.2024.2387609. [9]     D. H. Kim *et al.*, “Usefulness of Automatic Speech Recognition As-sessment of Children With Speech Sound Disorders: Validation Study,” Journal of Medical Internet Research, vol. 27, Art. no. e60520, 2025, doi: 10.2196/60520. [10]    A. Rustagi, S. Bajpai, N. Kaur, and S. Siddharth, “Dhvani: A Weakly-Supervised Phonemic Error Detection and Personalized Feedback Sys-tem for Hindi,” arXiv preprint arXiv:2506.02166, 2025.