TARQ: Tail-Aware Reconstruction Quantization for Rare-Word Robust Automatic Speech Recognition

arXiv:2605.27808v1 Announce Type: new Abstract: Data-aware post-training quantization (PTQ) minimizes a per-token reconstruction loss on a small calibration corpus, implicitly weighting positions by their empirical frequency. For extbf{A}utomatic extbf{S}peech extbf{R}ecognition (ASR), this misaligns with tail-sensitive risk: names, numerals, and domain-specific words receive proportionally little calibration mass. We propose extbf{Tail-Aware Reconstruction Quantization} (TARQ), a label-free PTQ framework that shifts calibration toward the lexical tail via extbf{rareBAL}, a closed-form per-Linear-layer rule equalizing common/tail mass, paired with a metric-consistent residual correction. TARQ requires no entity labels, no curated calibration set, no validation decoding, and no additional training. Across eight ASR backbones and six datasets at W4G128, TARQ improves mean rare-extbf{W}ord extbf{E}rror extbf{R}ate (rare-WER) without an aggregate-WER regression, achieves the lowest cross-corpus rare-WER swing among compared methods, and transfers to entity-rich benchmarks (ProfASR, ContextASR-Speech-En) without entity supervision.

Source: arXiv cs.CL | 2026-05-28