Transformer-Based Multi-Aspect Multi-Granularity Non-Native English Speaker Pronunciation Assessment
This work addresses the need for comprehensive pronunciation feedback for self-directed language learners, though it appears incremental as it builds on existing methods for a specific domain.
The paper tackles the problem of automatic pronunciation assessment by modeling multiple aspects (accuracy, fluency, completeness, prosody) at multiple granularities, achieving state-of-the-art results on the speechocean762 dataset using a Transformer-based model.
Automatic pronunciation assessment is an important technology to help self-directed language learners. While pronunciation quality has multiple aspects including accuracy, fluency, completeness, and prosody, previous efforts typically only model one aspect (e.g., accuracy) at one granularity (e.g., at the phoneme-level). In this work, we explore modeling multi-aspect pronunciation assessment at multiple granularities. Specifically, we train a Goodness Of Pronunciation feature-based Transformer (GOPT) with multi-task learning. Experiments show that GOPT achieves the best results on speechocean762 with a public automatic speech recognition (ASR) acoustic model trained on Librispeech.