DTM-Codec achieves better reconstruction quality and intelligibility than fixed-frame-rate neural speech codecs at matched total bitrate via dynamic token masking and Path Length Equalization for variable frame rates.
TaDi- Codec: Text-aware diffusion speech tokenizer for speech language modeling,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DTM-Codec: Dynamic Token Masking for VFR Speech Coding with Efficient Boundary Selection
DTM-Codec achieves better reconstruction quality and intelligibility than fixed-frame-rate neural speech codecs at matched total bitrate via dynamic token masking and Path Length Equalization for variable frame rates.