Audio-Reasoner: Improving Reasoning Capability in Large Audio Language Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916958593613824 |
|---|---|
| author | Xie, Zhifei Lin, Mingbao Liu, Zihang Wu, Pengcheng Yan, Shuicheng Miao, Chunyan |
| author_facet | Xie, Zhifei Lin, Mingbao Liu, Zihang Wu, Pengcheng Yan, Shuicheng Miao, Chunyan |
| contents | Recent advancements in multimodal reasoning have largely overlooked the audio modality. We introduce Audio-Reasoner, a large-scale audio language model for deep reasoning in audio tasks. We meticulously curated a large-scale and diverse multi-task audio dataset with simple annotations. Then, we leverage closed-source models to conduct secondary labeling, QA generation, along with structured COT process. These datasets together form a high-quality reasoning dataset with 1.2 million reasoning-rich samples, which we name CoTA. Following inference scaling principles, we train Audio-Reasoner on CoTA, enabling it to achieve great logical capabilities in audio reasoning. Experiments show state-of-the-art performance across key benchmarks, including MMAU-mini (+25.42%), AIR-Bench chat/foundation(+14.57%/+10.13%), and MELD (+8.01%). Our findings stress the core of structured CoT training in advancing audio reasoning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_02318 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Audio-Reasoner: Improving Reasoning Capability in Large Audio Language Models Xie, Zhifei Lin, Mingbao Liu, Zihang Wu, Pengcheng Yan, Shuicheng Miao, Chunyan Sound Artificial Intelligence Computation and Language Machine Learning Multimedia Audio and Speech Processing Recent advancements in multimodal reasoning have largely overlooked the audio modality. We introduce Audio-Reasoner, a large-scale audio language model for deep reasoning in audio tasks. We meticulously curated a large-scale and diverse multi-task audio dataset with simple annotations. Then, we leverage closed-source models to conduct secondary labeling, QA generation, along with structured COT process. These datasets together form a high-quality reasoning dataset with 1.2 million reasoning-rich samples, which we name CoTA. Following inference scaling principles, we train Audio-Reasoner on CoTA, enabling it to achieve great logical capabilities in audio reasoning. Experiments show state-of-the-art performance across key benchmarks, including MMAU-mini (+25.42%), AIR-Bench chat/foundation(+14.57%/+10.13%), and MELD (+8.01%). Our findings stress the core of structured CoT training in advancing audio reasoning. |
| title | Audio-Reasoner: Improving Reasoning Capability in Large Audio Language Models |
| topic | Sound Artificial Intelligence Computation and Language Machine Learning Multimedia Audio and Speech Processing |
| url | https://arxiv.org/abs/2503.02318 |