Japanese learners of English as a foreign language (EFL) often struggle with the pronunciation of specific American English (AE) vowels, particularly the mid and low vowels /æ/, /ɑ/, /ʌ/, /ɔ/, and /ɝ/, which tend to be perceptually assimilated into native vowel categories. This study explores the impact of a five-week, technology-enhanced vowel space training program that integrates dual-mode input—auditory and visual—to facilitate more accurate second language (L2) vowel acquisition. The participants were first-year Japanese undergraduates at a private university in Tokyo. The experimental group (N=30) received targeted training via an interactive, vowel-space mapping tool that employs dynamic color coding to represent vowel quality, providing multimodal reinforcement. In contrast, the control group (N=30) underwent conventional monomodal auditory training. To evaluate learning outcomes, both groups completed pre- and post-training production and identification tasks. The study’s findings are analyzed in the context of CALL-based phonetic training, addressing the pedagogical potential of dual-modality input in reshaping L2 vowel identification and production. Implications for technology-mediated pronunciation training and its role in optimizing L2 phonemic awareness and phonological development in EFL learners are discussed.