# Can large audio language models understand child stuttering speech? speech summarization, and source separation

arXiv: https://arxiv.org/abs/2510.20850 · HTML: https://mlsanity.com/paper/2510.20850/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 19 of its 31 references that are in the map.
- First posted on arXiv 21 Oct 2025.

## Builds on (most cited first, 19 of 19)

- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [Conformer: Convolution-augmented Transformer for Speech Recognition](https://mlsanity.com/paper/2005.08100/) (2020)
- [Qwen2-Audio Technical Report](https://mlsanity.com/paper/2407.10759/) (2024)
- [SALMONN: Towards Generic Hearing Abilities for Large Language Models](https://mlsanity.com/paper/2310.13289/) (2023)
- [Conv-TasNet: Surpassing Ideal Time-Frequency Magnitude Masking for Speech Separation](https://mlsanity.com/paper/1809.07454/) (2018)
- [MMAU: A Massive Multi-Task Audio Understanding and Reasoning Benchmark](https://mlsanity.com/paper/2410.19168/) (2024)
- [AudioGPT: Understanding and Generating Speech, Music, Sound, and Talking Head](https://mlsanity.com/paper/2304.12995/) (2023)
- [Transformer Transducer: A Streamable Speech Recognition Model with Transformer Encoders and RNN-T Loss](https://mlsanity.com/paper/2002.02562/) (2020)
- [AIR-Bench: Benchmarking Large Audio-Language Models via Generative Comprehension](https://mlsanity.com/paper/2402.07729/) (2024)
- [Audio Flamingo 3: Advancing Audio Intelligence with Fully Open Large Audio Language Models](https://mlsanity.com/paper/2507.08128/) (2025)
- [A Survey on Speech Large Language Models for Understanding](https://mlsanity.com/paper/2410.18908/) (2024)
- [Almost Unsupervised Text to Speech and Automatic Speech Recognition](https://mlsanity.com/paper/1905.06791/) (2019)
- [Towards Holistic Evaluation of Large Audio-Language Models: A Comprehensive Survey](https://mlsanity.com/paper/2505.15957/) (2025)
- [Towards Better Domain Adaptation for Self-supervised Models: A Case Study of Child ASR](https://mlsanity.com/paper/2305.00115/) (2023)
- [Afrispeech-Dialog: A Benchmark Dataset for Spontaneous English Conversations in Healthcare and Beyond](https://mlsanity.com/paper/2502.03945/) (2025)
- [Can Large Audio Language Models Understand Audio Well? Speech, Scene and Events Understanding Benchmark for LALMs](https://mlsanity.com/paper/2509.13148/) (2025)
- [State-Space Large Audio Language Models](https://mlsanity.com/paper/2411.15685/) (2024)
- [Domain-Aware Speaker Diarization On African-Accented English](https://mlsanity.com/paper/2509.21554/) (2025)
- [CHSER: A Dataset and Case Study on Generative Speech Error Correction for Child ASR](https://mlsanity.com/paper/2505.18463/) (2025)

## Built on by (most cited first, 2 of 2)

- [Reasoning Beyond Transcription: Audio Language Models on Child Stuttering Speech](https://mlsanity.com/paper/2609.07968/) (2026)
- [Role-guided Speaker Deletion Verification in Clinical Psychiatry Speech Recordings with Audio Language Models](https://mlsanity.com/paper/2609.38491/) (2026)

## Abstract (arXiv)

Child speech differs from adult speech in acoustics, prosody, and language development, and disfluencies (repetitions, prolongations, blocks) further challenge Automatic Speech Recognition (ASR) and downstream Natural Language Processing (NLP). Recent large audio-language models (LALMs) demonstrate strong cross-modal audio understanding; however, their behavior in disfluent child speech remains underexplored. We evaluate several state-of-the-art LALMs in two settings: an interview (mixed speakers) and a reading task (single child). The tasks are (i) single-channel source separation to isolate the child and (ii) child-only summarization that preserves clinically relevant disfluencies and avoids adult-speech leakage. Evaluation combines Large Language Model (LLM) as a judge, human expert ratings, and BERTScore (F1), and we report agreement between models and between models and humans to assess reliability. Our findings delineate the conditions under which LALMs produce faithful child-only summaries from mixed audio and where they fail, offering practical guidance for clinical and educational deployments. We provide prompts and evaluation scripts to support replication.