Adapts DAAM cross-attention attribution to speech diffusion models for the first time, mapping how style-caption tokens shape acoustic output in CapSpeech-TTS across layers and diffusion steps.
Cap- Speech: Enabling downstream applications in style-captioned text-to-speech,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
How Do Instructions Shape Speech? Cross-Attention Attribution for Style-Captioned Text-to-Speech
Adapts DAAM cross-attention attribution to speech diffusion models for the first time, mapping how style-caption tokens shape acoustic output in CapSpeech-TTS across layers and diffusion steps.