New Mamba-based model targets phoneme-level alignment in silent talking-face video
A new arXiv paper, Speaking-Clock Mamba, proposes phoneme-conditioned state-space modeling for visual forced alignment, locating phoneme and word boundaries in silent talking-face video given a transcript.
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-11-17 05:00 · arXiv cs.CV
Speaking-Clock Mamba: Phoneme-Conditioned State-Space Modeling for Visual Forced Alignment