Group Relative Policy Optimisation Improves Multilingual Speech Recognition in Low-Resource Languages

arXiv cs.AIOctober 23, 2026en

arXiv:2610.30663v1 Announce Type: new Abstract: Autoregressive encoder-decoder speech recognition models trained with teacher forcing can suffer from exposure bias, limiting their ability to recover from their own prediction errors at inference time. Recent work has shown that reinforcement learning with Group Relative Policy Optimisation (GRPO) improves speech recognition in major languages with LLM-based models. We extend this line of work to smaller, non-LLM encoder-decoder ASR models, using Whisper-large-v3 as a case study across 51 African languages. The benefits of this approach are most pronounced when training data is scarce or when input audio is noisy or degraded. Our resulting mod

This is a short summary published by AI Global Wire. The full article is owned and hosted by arXiv cs.AI — open it there to read it in full.

• Röst-AI
• Verktyg
• Forskning
• Reglering