Look, Listen, and Answer: Overcoming Biases for Audio-Visual Question Answering
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913718922641408 |
|---|---|
| author | Ma, Jie Hu, Min Wang, Pinghui Sun, Wangchun Song, Lingyun Pei, Hongbin Liu, Jun Du, Youtian |
| author_facet | Ma, Jie Hu, Min Wang, Pinghui Sun, Wangchun Song, Lingyun Pei, Hongbin Liu, Jun Du, Youtian |
| contents | Audio-Visual Question Answering (AVQA) is a complex multi-modal reasoning task, demanding intelligent systems to accurately respond to natural language queries based on audio-video input pairs. Nevertheless, prevalent AVQA approaches are prone to overlearning dataset biases, resulting in poor robustness. Furthermore, current datasets may not provide a precise diagnostic for these methods. To tackle these challenges, firstly, we propose a novel dataset, MUSIC-AVQA-R, crafted in two steps: rephrasing questions within the test split of a public dataset (MUSIC-AVQA) and subsequently introducing distribution shifts to split questions. The former leads to a large, diverse test space, while the latter results in a comprehensive robustness evaluation on rare, frequent, and overall questions. Secondly, we propose a robust architecture that utilizes a multifaceted cycle collaborative debiasing strategy to overcome bias learning. Experimental results show that this architecture achieves state-of-the-art performance on MUSIC-AVQA-R, notably obtaining a significant improvement of 9.32%. Extensive ablation experiments are conducted on the two datasets mentioned to analyze the component effectiveness within the debiasing strategy. Additionally, we highlight the limited robustness of existing multi-modal QA methods through the evaluation on our dataset. We also conduct experiments combining various baselines with our proposed strategy on two datasets to verify its plug-and-play capability. Our dataset and code are available at https://github.com/reml-group/MUSIC-AVQA-R. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_12020 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Look, Listen, and Answer: Overcoming Biases for Audio-Visual Question Answering Ma, Jie Hu, Min Wang, Pinghui Sun, Wangchun Song, Lingyun Pei, Hongbin Liu, Jun Du, Youtian Computer Vision and Pattern Recognition I.2.10 Audio-Visual Question Answering (AVQA) is a complex multi-modal reasoning task, demanding intelligent systems to accurately respond to natural language queries based on audio-video input pairs. Nevertheless, prevalent AVQA approaches are prone to overlearning dataset biases, resulting in poor robustness. Furthermore, current datasets may not provide a precise diagnostic for these methods. To tackle these challenges, firstly, we propose a novel dataset, MUSIC-AVQA-R, crafted in two steps: rephrasing questions within the test split of a public dataset (MUSIC-AVQA) and subsequently introducing distribution shifts to split questions. The former leads to a large, diverse test space, while the latter results in a comprehensive robustness evaluation on rare, frequent, and overall questions. Secondly, we propose a robust architecture that utilizes a multifaceted cycle collaborative debiasing strategy to overcome bias learning. Experimental results show that this architecture achieves state-of-the-art performance on MUSIC-AVQA-R, notably obtaining a significant improvement of 9.32%. Extensive ablation experiments are conducted on the two datasets mentioned to analyze the component effectiveness within the debiasing strategy. Additionally, we highlight the limited robustness of existing multi-modal QA methods through the evaluation on our dataset. We also conduct experiments combining various baselines with our proposed strategy on two datasets to verify its plug-and-play capability. Our dataset and code are available at https://github.com/reml-group/MUSIC-AVQA-R. |
| title | Look, Listen, and Answer: Overcoming Biases for Audio-Visual Question Answering |
| topic | Computer Vision and Pattern Recognition I.2.10 |
| url | https://arxiv.org/abs/2404.12020 |