The year/Independent research

Paper 2605.19833

Mega-ASR: Towards In-the-wild^2 Speech Recognition via Scaling up Real-world Acoustic Simulation

Published
May 2026
Research lab
Independent
Citations
1
GitHub
1.1K stars

01 In brief

Summary

MEGA-ASR is a framework for automatic speech recognition (ASR) in real-world environments, addressing the 'acoustic robustness bottleneck' where models fail under severe, compositional distortions.

The authors introduce VOICES-IN-THE-WILD-2M, a large-scale dataset with 7 atomic acoustic effects (noise, far-field, obstructed, echo&reverb, recording, electronic distortion, transmission dropout) and 54 physically plausible compound scenarios, built via spectral-manipulation simulation.

Training uses Acoustic-to-Semantic Progressive Supervised Fine-Tuning (A2S-SFT) and Dual-Granularity WER-Gated Policy Optimization (DG-WGPO), which combines token-level and sentence-level rewards with WER-gated fusion.

Experiments show MEGA-ASR outperforms prior systems on adverse-condition benchmarks (e.g., 45.69% vs.

54.01% WER on VOiCES R4-B-F, 21.49% vs.

29.34% on NOIZEUS Sta-0) and achieves over 30% relative WER reduction on complex compositional scenarios.

An environment-aware router preserves clean-speech performance by routing inputs to either the robust or base model.

The work establishes a scalable paradigm for robust ASR in-the-wild, with data and benchmarks publicly released.

02 From the paper

Abstract

Despite rapid advances in automatic speech recognition (ASR) and large audio-language models, robust recognition in real-world environments remains limited by an "acoustic robustness bottleneck": models often lose acoustic grounding and produce omissions or hallucinations under severe, compositional distortions. We propose Mega-ASR, a unified ASR-in-the-wild framework that combines scalable compound-data construction with progressive acoustic-to-semantic optimization. We introduce Voices-in-the-Wild-2M, covering 7 classic acoustic phenomena and 54 physically plausible compound scenarios, and train Mega-ASR with Acoustic-to-Semantic Progressive Supervised Fine-Tuning and Dual-Granularity WER-Gated Policy Optimization. Extensive experiments demonstrate that Mega-ASR achieves significant advantages over prior state-of-the-art systems on adverse-condition ASR benchmarks (45.69% vs. 54.01% on VOiCES R4-B-F, and 21.49% vs. 29.34% on NOIZEUS Sta-0). On complex compositional acoustic scenarios, Mega-ASR further delivers over 30% relative WER reduction against strong open- and closed-source baselines, establishing a scalable paradigm for robust ASR in-the-wild.