基于早期深度特征融合的鸟类声学分类

Q4 Agricultural and Biological Sciences Western Birds Pub Date : 2023-03-01 DOI:10.3390/birds4010011
Jie Xie, Mingying Zhu
{"title":"基于早期深度特征融合的鸟类声学分类","authors":"Jie Xie, Mingying Zhu","doi":"10.3390/birds4010011","DOIUrl":null,"url":null,"abstract":"Bird sound classification plays an important role in large-scale temporal and spatial environmental monitoring. In this paper, we investigate both transfer learning and training from scratch for bird sound classification, where pre-trained models are used as feature extractors. Specifically, deep cascade features are extracted from various layers of different pre-trained models, which are then fused to classify bird sounds. A multi-view spectrogram is constructed to characterize bird sounds by simply repeating the spectrogram to make it suitable for pre-trained models. Furthermore, both mixup and pitch shift are applied for augmenting bird sounds to improve the classification performance. Experimental classification on 43 bird species using linear SVM indicates that deep cascade features can achieve the highest balanced accuracy of 90.94% ± 1.53%. To further improve the classification performance, an early fusion method is used by combining deep cascaded features extracted from different pre-trained models. The final best classification balanced accuracy is 94.89% ± 1.35%.","PeriodicalId":52426,"journal":{"name":"Western Birds","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Acoustic Classification of Bird Species Using an Early Fusion of Deep Features\",\"authors\":\"Jie Xie, Mingying Zhu\",\"doi\":\"10.3390/birds4010011\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Bird sound classification plays an important role in large-scale temporal and spatial environmental monitoring. In this paper, we investigate both transfer learning and training from scratch for bird sound classification, where pre-trained models are used as feature extractors. Specifically, deep cascade features are extracted from various layers of different pre-trained models, which are then fused to classify bird sounds. A multi-view spectrogram is constructed to characterize bird sounds by simply repeating the spectrogram to make it suitable for pre-trained models. Furthermore, both mixup and pitch shift are applied for augmenting bird sounds to improve the classification performance. Experimental classification on 43 bird species using linear SVM indicates that deep cascade features can achieve the highest balanced accuracy of 90.94% ± 1.53%. To further improve the classification performance, an early fusion method is used by combining deep cascaded features extracted from different pre-trained models. The final best classification balanced accuracy is 94.89% ± 1.35%.\",\"PeriodicalId\":52426,\"journal\":{\"name\":\"Western Birds\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-03-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Western Birds\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.3390/birds4010011\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"Agricultural and Biological Sciences\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Western Birds","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3390/birds4010011","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"Agricultural and Biological Sciences","Score":null,"Total":0}
引用次数: 1

摘要

鸟声分类在大尺度时空环境监测中具有重要作用。在本文中,我们研究了迁移学习和从头开始训练用于鸟类声音分类,其中使用预训练模型作为特征提取器。具体来说,从不同预训练模型的各个层中提取深度级联特征,然后将其融合到鸟类声音分类中。通过简单地重复声谱图,构建了一个多视图声谱图来表征鸟类的声音,使其适合于预训练的模型。在此基础上,利用混频和移频两种方法增强了鸟类的叫声,提高了分类性能。利用线性支持向量机对43种鸟类进行分类的实验表明,深度级联特征可以达到最高的平衡准确率(90.94%±1.53%)。为了进一步提高分类性能,采用一种早期融合方法,将从不同预训练模型中提取的深度级联特征组合在一起。最终的最佳分类平衡准确率为94.89%±1.35%。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Acoustic Classification of Bird Species Using an Early Fusion of Deep Features
Bird sound classification plays an important role in large-scale temporal and spatial environmental monitoring. In this paper, we investigate both transfer learning and training from scratch for bird sound classification, where pre-trained models are used as feature extractors. Specifically, deep cascade features are extracted from various layers of different pre-trained models, which are then fused to classify bird sounds. A multi-view spectrogram is constructed to characterize bird sounds by simply repeating the spectrogram to make it suitable for pre-trained models. Furthermore, both mixup and pitch shift are applied for augmenting bird sounds to improve the classification performance. Experimental classification on 43 bird species using linear SVM indicates that deep cascade features can achieve the highest balanced accuracy of 90.94% ± 1.53%. To further improve the classification performance, an early fusion method is used by combining deep cascaded features extracted from different pre-trained models. The final best classification balanced accuracy is 94.89% ± 1.35%.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Western Birds
Western Birds Agricultural and Biological Sciences-Animal Science and Zoology
CiteScore
0.60
自引率
0.00%
发文量
0
期刊最新文献
David F. DeSante’s Birds of Cabo San Lucas, Fall 1968: A Historic Account First Record of Tricolored Blackbirds in Idaho American Crow Cracks Open Bivalve via Automobile Second Prebasic Molt of a Black-headed Gull at Anchorage, Alaska Nesting Bald Eagle Population Numbers, Density, Territorial Resources, and Relationship to Human Development in Northern Colorado’s Front Range
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1