arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2512.03563cs.SDcs.AI
收藏

生物声学中的状态空间模型:与Transformer的比较评估

State Space Models for Bioacoustics: A Comparative Evaluation with Transformers

Auburn University(阿伯茨福德大学)

Chengyu Tang, Sanjeev Baskiyar

AI总结:

本文评估了基于Mamba架构的BioMamba模型在生物声学中的表现,展示了其在减少VRAM消耗的同时与Transformer模型相媲美。

摘要:

在本研究中,我们通过引入BioMamba,一种基于Mamba的音频表示模型,评估了Mamba架构在生物声学中的有效性。我们使用自监督学习在大型音频语料库上预训练BioMamba,并在BEANS基准上评估其在多样化的分类和检测任务中的性能。与最先进的基于Transformer的模型(AVES)相比,BioMamba在性能上相当,但显著减少了VRAM的消耗。我们的结果表明,Mamba作为一种计算效率高的替代方案,在实际的环境监测中具有潜力。

英文摘要:

In this study, we evaluate the efficacy of the Mamba architecture bioacoustics by introducing BioMamba, a Mamba-based audio representation model for wildlife sounds. We pre-train a BioMamba using self-supervised learning on a large audio corpus and evaluate it on the BEANS benchmark across diverse classification and detection tasks. Compared to the state-of-the-art Transformer-based model (AVES), BioMamba achieves comparable performance while significantly reducing VRAM consumption. Our results demonstrate Mamba's potential as a computationally efficient alternative for real-world environmental monitoring.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑