AVERT: Audio-Verified Adjudication for Spoken Dialogue State Tracking

Researchers introduced AVERT, an audio-verified adjudication method for spoken dialogue state tracking that combines cross-turn agreement with a trained audio-conditioned verifier, lifting JGA on SpokenWOZ from 38.34 with a text editor alone to 40.13 without retraining the base speech-LLM or editor. The method resolves three error types—inconsistent values, omitted slots, and audio-unsupported values—using vote, add, and swap operators restricted to specific slots, with the audio verifier providing a statistically significant gain.

arXiv:2609.01828v1 Announce Type: new Abstract: Spoken dialogue state tracking recovers slot-value pairs from speech, where ASR errors concentrate in entity values and persist across turns, making it both a generation and an editing problem. A strong per-turn text editor corrects much of this but, operating on the transcript alone, leaves three recoverable errors: a value predicted inconsistently across turns, an omitted slot, and a value the audio does not support. We present AVERT, which scores each candidate value by combining cross-turn agreement with a trained audio-conditioned verifier and resolves the three error types with three operators, vote, add, and swap, each restricted to the slots where its error is common. On SpokenWOZ, a base speech-LLM reaches 33.04 JGA, a text editor 38.34, and AVERT 40.13, without retraining either. This is in the range of a 1B end-to-end system that consumes the full spoken history 39.32 , though AVERT uses two 1B decoders rather than one. The audio verifier contributes a statistically significant gain, and restricting each operator to a selected slot subset matters: removing it lets unrestricted voting overwrite correct categorical values and fall below the editor.