SOURCE-LINKED INTELLIGENCE

GrainSpeech: Less Context, More Detail for Compact Speech Synthesis

arXiv · AI, language, vision and robotics · article · Sep 16, 2026 · UTC

Compact acoustic models face a challenging quality-capacity trade-off. We investigate two factors in this regime: encoder context and Mel-spectrogram supervision. A receptive-field-scaling study shows that expanding self-attention beyond 15 phonemes provides no consistent gains in pitch, energy, or duration prediction. Guided by this finding, we introduce a fixed-receptive-field convolutional encoder that reduces the respective prediction errors by 36.0%, 17.3%, and 3.4%. We further show that directly transferring image-domain gradient-variance supervision restores fine-scale variation but deg

Read original source ↗ Open in workspace

recordType

paper

region

Global