AQAScore: Evaluating Semantic Alignment in Text-to-Audio Generation via Audio Question Answering
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912837181374464 |
|---|---|
| author | Kuan, Chun-Yi Chang, Kai-Wei Lee, Hung-yi |
| author_facet | Kuan, Chun-Yi Chang, Kai-Wei Lee, Hung-yi |
| contents | Although text-to-audio generation has made remarkable progress in realism and diversity, the development of evaluation metrics has not kept pace. Widely-adopted approaches, typically based on embedding similarity like CLAPScore, effectively measure general relevance but remain limited in fine-grained semantic alignment and compositional reasoning. To address this, we introduce AQAScore, a backbone-agnostic evaluation framework that leverages the reasoning capabilities of audio-aware large language models (ALLMs). AQAScore reformulates assessment as a probabilistic semantic verification task; rather than relying on open-ended text generation, it estimates alignment by computing the exact log-probability of a "Yes" answer to targeted semantic queries. We evaluate AQAScore across multiple benchmarks, including human-rated relevance, pairwise comparison, and compositional reasoning tasks. Experimental results show that AQAScore consistently achieves higher correlation with human judgments than similarity-based metrics and generative prompting baselines, showing its effectiveness in capturing subtle semantic inconsistencies and scaling with the capability of underlying ALLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_14728 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AQAScore: Evaluating Semantic Alignment in Text-to-Audio Generation via Audio Question Answering Kuan, Chun-Yi Chang, Kai-Wei Lee, Hung-yi Audio and Speech Processing Artificial Intelligence Computation and Language Machine Learning Sound Although text-to-audio generation has made remarkable progress in realism and diversity, the development of evaluation metrics has not kept pace. Widely-adopted approaches, typically based on embedding similarity like CLAPScore, effectively measure general relevance but remain limited in fine-grained semantic alignment and compositional reasoning. To address this, we introduce AQAScore, a backbone-agnostic evaluation framework that leverages the reasoning capabilities of audio-aware large language models (ALLMs). AQAScore reformulates assessment as a probabilistic semantic verification task; rather than relying on open-ended text generation, it estimates alignment by computing the exact log-probability of a "Yes" answer to targeted semantic queries. We evaluate AQAScore across multiple benchmarks, including human-rated relevance, pairwise comparison, and compositional reasoning tasks. Experimental results show that AQAScore consistently achieves higher correlation with human judgments than similarity-based metrics and generative prompting baselines, showing its effectiveness in capturing subtle semantic inconsistencies and scaling with the capability of underlying ALLMs. |
| title | AQAScore: Evaluating Semantic Alignment in Text-to-Audio Generation via Audio Question Answering |
| topic | Audio and Speech Processing Artificial Intelligence Computation and Language Machine Learning Sound |
| url | https://arxiv.org/abs/2601.14728 |