Estimating articulatory movements in speech production with transformer networks
This work addresses a domain-specific challenge in speech processing by improving alignment and estimation accuracy for articulatory movements, though it appears incremental as it builds on existing transformer methods.
The paper tackles the problem of estimating articulatory movements in speech production from acoustics and phonemes, achieving relative improvements of up to 154% in correlation coefficient for phoneme-to-articulatory estimation and up to 3.1% for acoustic-to-articulatory inversion compared to state-of-the-art baselines.
We estimate articulatory movements in speech production from different modalities - acoustics and phonemes. Acoustic-to articulatory inversion (AAI) is a sequence-to-sequence task. On the other hand, phoneme to articulatory (PTA) motion estimation faces a key challenge in reliably aligning the text and the articulatory movements. To address this challenge, we explore the use of a transformer architecture - FastSpeech, with explicit duration modelling to learn hard alignments between the phonemes and articulatory movements. We also train a transformer model on AAI. We use correlation coefficient (CC) and root mean squared error (rMSE) to assess the estimation performance in comparison to existing methods on both tasks. We observe 154%, 11.8% & 4.8% relative improvement in CC with subject-dependent, pooled and fine-tuning strategies, respectively, for PTA estimation. Additionally, on the AAI task, we obtain 1.5%, 3% and 3.1% relative gain in CC on the same setups compared to the state-of-the-art baseline. We further present the computational benefits of having transformer architecture as representation blocks.