对比声音令牌与传统卷积模型在鸟类叫声检测中的表现。
Can Tokens Compete? Token Representations against Supervised CNN Backbones for BirdCLEF+ 2026
- 用冻结的Perch v2和HGNetV2-B0构建监督基线。
- 在90分钟内实现0.936的私榜分数，排名1894。
- 首次系统比较音频编码器与AudioSet训练的令牌模型。
本文介绍DS@GT ARC团队在BirdCLEF+ 2026中的方法，任务为从潘塔纳尔湿地声景中进行多标签动物叫声检测。2026版新增约一小时标注数据，使任务更适配监督式流水线。首先，构建一个集成模型，包含冻结的Perch v2主干网络、训练好的HGNetV2-B0声事件检测网络及非鸟原型头，在90分钟CPU预算内取得0.936的私榜分数，排名1894。其次，探讨令牌表示能否与监督卷积主干竞争，对比神经音频编码器生成的码本表示与基础嵌入生成的语义表示。将两个生物声学专用模型与四个在AudioSet上预训练的令牌编码器进行对比。代码仓库见https://github.com/dsgt-arc/birdclef-2026。
原文摘要 · Abstract (English)
This paper details the DS@GT ARC team's approach to BirdCLEF+ 2026, multi-label detection of animal vocalizations in soundscapes from the Pantanal wetlands. The 2026 edition adds about an hour of labeled soundscapes, shifting the task toward supervised pipelines fit to the labeled set. First, we build a competitive supervised baseline that ensembles a frozen Perch v2 backbone, a trained HGNetV2-B0 sound-event-detection network, and a non-bird prototypical head, reaching a private leaderboard score of 0.936 at rank 1894 within a 90-minute CPU budget. Second, we ask whether token-based representations can compete, contrasting codec representations from neural audio codecs against semantic representations from foundational embeddings. We compare two bioacoustic specialist models against four token-based encoders trained on AudioSet. The repository for this work can be found at https://github.com/dsgt-arc/birdclef-2026.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。