Gradient saliency of teacher-forced token log-probabilities, decoded by dynamic programming, yields usable word alignments for every ASR family including speech LLMs.
Attention-Constrained Inference for Robust Decoder-Only Text-to-Speech
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Recent popular decoder-only text-to-speech models are known for their ability of generating natural-sounding speech. However, such models sometimes suffer from word skipping and repeating due to the lack of explicit monotonic alignment constraints. In this paper, we notice from the attention maps that some particular attention heads of the decoder-only model indicate the alignments between speech and text. We call the attention maps of those heads Alignment-Emerged Attention Maps (AEAMs). Based on this discovery, we propose a novel inference method without altering the training process, named Attention-Constrained Inference (ACI), to facilitate monotonic synthesis. It first identifies AEAMs using the Attention Sweeping algorithm and then applies constraining masks on AEAMs. Our experimental results on decoder-only TTS model VALL-E show that the WER of synthesized speech is reduced by up to 20.5% relatively with ACI while the naturalness and speaker similarity are comparable.
fields
cs.CL 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Gradient-Based Speech-to-Text Alignment for Any ASR Model: From CTC to Speech LLMs
Gradient saliency of teacher-forced token log-probabilities, decoded by dynamic programming, yields usable word alignments for every ASR family including speech LLMs.