Voices of Civilizations: A Multilingual QA Benchmark for Global Music Understanding
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911475317080064 |
|---|---|
| author | Wu, Shangda Zhou, Ziya Zang, Yongyi Zheng, Yutong Liang, Dafang Yuan, Ruibin Kong, Qiuqiang |
| author_facet | Wu, Shangda Zhou, Ziya Zang, Yongyi Zheng, Yutong Liang, Dafang Yuan, Ruibin Kong, Qiuqiang |
| contents | We introduce Voices of Civilizations, the first multilingual QA benchmark for evaluating audio LLMs' cultural comprehension on full-length music recordings. Covering 380 tracks across 38 languages, our automated pipeline yields 1,190 multiple-choice questions through four stages - each followed by manual verification: 1) compiling a representative music list; 2) generating cultural-background documents for each sample in the music list via LLMs; 3) extracting key attributes from those documents; and 4) constructing multiple-choice questions probing language, region associations, mood, and thematic content. We evaluate models under four conditions and report per-language accuracy. Our findings demonstrate that even state-of-the-art audio LLMs struggle to capture subtle cultural nuances without rich textual context and exhibit systematic biases in interpreting music from different cultural traditions. The dataset is publicly available on Hugging Face to foster culturally inclusive music understanding research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_00533 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Voices of Civilizations: A Multilingual QA Benchmark for Global Music Understanding Wu, Shangda Zhou, Ziya Zang, Yongyi Zheng, Yutong Liang, Dafang Yuan, Ruibin Kong, Qiuqiang Sound Audio and Speech Processing We introduce Voices of Civilizations, the first multilingual QA benchmark for evaluating audio LLMs' cultural comprehension on full-length music recordings. Covering 380 tracks across 38 languages, our automated pipeline yields 1,190 multiple-choice questions through four stages - each followed by manual verification: 1) compiling a representative music list; 2) generating cultural-background documents for each sample in the music list via LLMs; 3) extracting key attributes from those documents; and 4) constructing multiple-choice questions probing language, region associations, mood, and thematic content. We evaluate models under four conditions and report per-language accuracy. Our findings demonstrate that even state-of-the-art audio LLMs struggle to capture subtle cultural nuances without rich textual context and exhibit systematic biases in interpreting music from different cultural traditions. The dataset is publicly available on Hugging Face to foster culturally inclusive music understanding research. |
| title | Voices of Civilizations: A Multilingual QA Benchmark for Global Music Understanding |
| topic | Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2603.00533 |