A Multi-Feature Fusion Bird Song Recognition Method Based on MobileViT

Abstract

Abstract: Objective: Monitoring bird diversity is a major challenge, as birds are more easily heard rather than seen. To improve the
accuracy of bird song recognition, a multi-feature fusion bird song recognition method based on MobileViT is proposed. Methods:
Using the Beijing BirdsData database as the research object, three different spectrogram sample sets were extracted from preprocessed bird song signals and used as inputs to train three single feature models based on MobileViT. Finally, adaptive weighted feature fusion on the three single feature models was performed. Results: The recognition accuracy of the multi-feature fusion model on the spectrogram test set reached 97.57%, with an improvement of 1.77%-2.24% compared to the single feature model. Conclusion: Different spectrograms extracted from the same bird song signal exhibit differences in the characteristics of bird song, and multifeature fusion models can learn more extensive information from them, thereby significantly improving recognition accuracy.

Key words: bird song recognition, MobileViT, multi-feature fusion

CLC Number:

TP391

Luo Chuang, Chen Bobo, Zhang Jilei, Li Yanpeng, Fang Yihao, Su Xukun, Zhao Enming. A Multi-Feature Fusion Bird Song Recognition Method Based on MobileViT[J]. Journal of Dali University, 2025, 10(6): 38-46.

References

〔1〕 HUA F Y，WANG X Y，ZHENG X L，et al. Opportunities
for Biodiversity Gains under the World's Largest Reforesta⁃
tion Programme〔J〕. Nature Communications，2016，7（1）：
12717.
〔2〕 LOVETTE I J，FITZPATRICK J W. Handbook of Bird Biol⁃
ogy〔M〕. Hoboken：Wiley-Blackwell，2016.
〔3〕孙戈，曾立雄，钱法文，等. 鸟类监测技术现状与发展趋
势〔J〕. 地理信息世界，2022，29（6）：26-29.
〔4〕张明明，梁宇祥. 简析鸟类的野外调查和监测技术〔J〕.
林业科技，2023，48（6）：41-44.
〔5〕 WOOD C M，GUTIÉRREZ R J，ZACHARIAH PEERY M.
Acoustic Monitoring Reveals a Diverse Forest Owl Commu⁃
nity，Illustrating Its Potential for Basic and Applied Ecology
〔J〕. Ecology，2019，100（9）：e02764.
〔6〕 ANDERSON S E，DAVE A S，MARGOLIASH D. Template-
Based Automatic Recognition of Birdsong Syllables from
Continuous Recordings〔J〕. The Journal of the Acoustical
Society of America，1996，100（2 Pt 1）：1209-1219.
〔7〕阙鑫华，乔倩，蒋慧，等. 基于改进DTW算法的海岛水鸟
鸣声识别应用研究〔J〕. 农村经济与科技，2021，32（11）：
320-322.
〔8〕 ACEVEDO M A，CORRADA-BRAVO C J，CORRADABRAVO
H，et al. Automated Classification of Bird and Am⁃
phibian Calls Using Machine Learning：A Comparison of
Methods〔J〕. Ecological Informatics，2009，4（4）：206-214.
〔9〕 SHIU Y，PALMER K J，ROCH M A，et al. Deep Neural
Networks for Automated Detection of Marine Mammal
Species〔J〕. Scientific Reports，2020，10（1）：607.
〔10〕谢卓钒，李鼎昭，孙海信，等. 面向鸟鸣声识别任务的深
度学习技术〔J〕. 生物多样性，2023，31（1）：137-144.
〔11〕 CARVALHO S，GOMES E F. Automatic Classification of
Bird Sounds：Using MFCC and Mel Spectrogram Features
with Deep Learning〔J〕. Vietnam Journal of Computer
Science，2023，10（1）：39-54.
〔12〕谢将剑，李文彬，张军国，等. 基于Chirplet语图特征和
深度学习的鸟类物种识别方法〔J〕. 北京林业大学学
报，2018，40（3）：122-127.
〔13〕HONG L. Acoustic Bird Species Recognition at BirdCLEF
2023：Training Strategies for Convolutional Neural Net⁃
work and Inference Acceleration Using OpenVINO〔C〕//
CLEF（ Working Notes）. 2023：2043-2050.
〔14〕 KAHL S，WOOD C M，EIBL M，et al. BirdNET：A Deep
Learning Solution for Avian Diversity Monitoring〔J〕. Eco⁃
logical Informatics，2021，61：101236.
〔15〕王基豪，周晓彦，李大鹏，等. 基于卷积神经网络和
Transformer 网络的鸟声识别〔J〕. 声学技术，2023，42
（5）：675-683.
〔16〕 MOLAU S，PITZ M，SCHLUTER R，et al. Computing
Mel-Frequency Cepstral Coefficients on the Power Spec⁃
trum〔C〕//2001 IEEE International Conference on Acous⁃
tics，Speech，and Signal Processing. Proceedings. May 7-
11，2001，Salt Lake City，UT，USA. 2001：73-76.
〔17〕 GLOTIN H，RICARD J，BALESTRIERO R. Fast Chirplet
Transform to Enhance CNN Machine Listening - Valida⁃
tion on Animal Calls and Speech〔EB/OL〕. 〔2024-03-
04〕.https：//arxiv.org/abs/1611.08749v2.
〔18〕 MEHTA S，RASTEGARI M. MobileViT：Light-Weight，
General-Purpose，and Mobile-Friendly Vision Trans⁃
forme〔r EB/OL〕.〔 2024-03-04〕. https：//arxiv.org/abs/2110.
02178v2.
〔19〕DOSOVITSKIY A，BEYER L，KOLESNIKOV A，et al. An
Image Is Worth 16x16 Words：Transformers for Image
Recognition at Scale〔EB/OL〕. 〔2024-03-04〕. https：//
arxiv.org/pdf/2010. 11929.
〔20〕马维娣，吴钦木. 基于改进轻量化网络MobileViT的苹
果树叶病害识别〔J〕. 江苏农业科学，2024，52（3）：229-
236.
〔21〕 CHENG Q Q，LI X H，ZHU B，et al. Drone Detection
Method Based on MobileViT and CA-PANet〔J〕. Elec⁃

tronics，2023，12（1）：223.

〔22〕周旷，姜名. 基于迁移学习的小样本目标识别研究进展
与展望〔J〕. 航空科学技术，2023，34（2）：1-9.
〔23〕随婷婷，王晓峰. 一种基于CLMF的深度卷积神经网络
模型〔J〕. 自动化学报，2016，42（6）：875-882.
〔24〕裴燚，刘光宇，雷远彬，等.基于RetinaFace 和FaceNet
算法的佩戴口罩人脸识别系统研究〔J〕. 大理大学学
报，2024，9（12）：51-57.
〔25〕赵恩铭，姚志强，杨松，等.基于机器视觉的自动目标
跟随系统设计与实现〔J〕.大理大学学报，2022，7（6）：
42-47.

[1]	Ma Xin, Wang Caikui, Hu Anshun, Liu Xuan, Fang Lianbiao, Luo Guilan. Modeling and Simulation Analysis of Regional Climate Complex Hierarchical Network in Yunnan Province [J]. Journal of Dali University, 2024, 9(12): 36-45.
[2]	Feng Dacheng, Zhang Lulu, Li Hui. Identity Authentication System Based on QR Code and Face Recognition [J]. Journal of Dali University, 2024, 9(12): 46-50.
[3]	Pei Yi, Liu Guangyu, Lei Yuanbin, Zhang Jilei, Chen Bobo, Zhao Hongyi, Zhao Enming. Masks-Wearing Face Recognition System Based on RetinaFace and FaceNet Algorithms [J]. Journal of Dali University, 2024, 9(12): 51-57.
[4]	Chen Bobo, Lei Yuanbin, Zhang Jilei, Zhao Hongyi, Luo Chuang, Zhao Enming. Cattle Posture Estimation Based on YOLO Algorithm [J]. Journal of Dali University, 2024, 9(12): 58-64.
[5]	Chen Bin . Research on Human Abnormal Emotion Recognition Algorithm Based on Video Images [J]. Journal of Dali University, 2023, 8(12): 27-31.
[6]	NiuYitong , Meng Deyao , Wu Ronghai, Ren Guopeng, Yang Dengqi. Comparative Study of Garbage Image Classification Models Based on Transfer Learning [J]. Journal of Dali University, 2022, 7(12): 8-14.
[7]	Huang Liusong, Zhang Fei, Zhang Xuefeng, Ｗu Huilin. Heterogeneous Database Security Synchronization System Based on GAP [J]. Journal of Dali University, 2022, 7(12): 25-30.
[8]	Zhao Enming, Yang Song, Luo Chuang, Liu Guangyu, Zhu Xiaoliang. Research on Construction Workers' Safety Protection Equipment Wearing Detection Based on YOLOv5s [J]. Journal of Dali University, 2022, 7(12): 37-42.
[9]	Zeng Xin, Zhang Xiaoling, Li Xiaowei. An Effective Algorithm for Mining High Average-Utility Co-Location Patterns [J]. Journal of Dali University, 2022, 7(6): 9-17.
[10]	Shi Chenyang, Yang Zizhong, Chen Xiangyu, Zhang Yuhang, Qian Zhengkun, Wang Yongliang, Wang Jianming. Research and Construction of Spider Recognition Platform Based on Transfer Learning and Data Augmentation [J]. Journal of Dali University, 2022, 7(6): 26-36.
[11]	Zeng Xin, Li Zhenpeng, Yang Jian, Li Xiaowei. High Average-Utility Co-Location Patterns Mining Algorithm with Rare Features [J]. Journal of Dali University, 2021, 6(12): 5-11.
[12]	Zhao Enming, Yang Song, Yao Zhiqiang. Design and Implementation of Ethnic Costume Recognition System Based on Faster-RCNN [J]. Journal of Dali University, 2021, 6(12): 23-28.
[13]	DING Li, CENG Shui-Ling. Research on Handwritten Hmong Character Recognition Based on Probabilistic Neural Network [J]. , 2018, 3(12): 18-23.
[14]	Luo Guilan1, Yang Zizhong2, Zhang Mei1, Wu Shilong1, Zhang Qian1, An Dong1. Insect Intelligent Identification and Real-time Monitoring System in Erhai Wetland [J]. , 2018, 3(6): 6-12.
[15]	Huang Jiaheng, Li Xiaowei, Chen Benhui, Yang Dengqi. A Comparative Study on Image Similarity Algorithms Based on Hash [J]. , 2017, 2(12): 32-37.