XU Xiajuan,LI Zhongcheng,XING Xiaoying

• 1:College of Wildlife and Protected Area
• 2: Northeast Forestry University
• 3:Northeast Asia Biodiversity Research Center

摘要(Abstract)：

Bird vocalizations are essential bioacoustic signals for species identification and ecological monitoring. The proliferation of acoustic monitoring technology has driven a paradigm shift from traditional machine learning to deep learning,which excels at extracting robust representations from noisy, large-scale field recordings. This paper systematically reviews the evolution and application of core deep learning architectures in avian bioacoustics from 2014 to 2025, including convolutional neural networks(CNNs), recurrent neural networks(RNNs), generative adversarial networks(GANs), and transformers. While CNNs dominate these domains due to their efficiency in processing spectrograms, transformer-based foundation models are increasingly reshaping the field. Despite notable progress, current research still faces several critical challenges, including the recognition of long-tail species, acoustic signal overlap in complex soundscapes, poor model generalization in out-of-distribution scenarios, and a fundamental lack of model interpretability. To address these limitations, future research should prioritize the construction of standardized, multi-dimensional datasets and the integration of advanced paradigms, specifically self-supervised learning(SSL), few-shot learning(FSL), and deep source separation. Furthermore, we emphasize the necessity of explainable artificial intelligence(XAI) and domain-specific ecological priors to bridge the gap between computational complexity and ethological transparency. By providing a concise synthesis of methods, applications, and research directions, this review aims to guide future work toward scalable, robust automated bird-sound analysis for conservation and ecological science and to support policy and management decisions globally and locally.

关键词(KeyWords)：

Abstract:

Keywords:

基金项目(Foundation): supported by the National Natural Science Foundation of China (32170485, 31501867)

作者(Author): XU Xiajuan,LI Zhongcheng,XING Xiaoying

参考文献(References)：

• [1]Pahuja R,Kumar A.Sound-spectrogram based automatic bird species recognition using MLP classifier.Appl.Acoust.,2021,180:108077
• [2]Ghani B,Hallerberg S.A randomized bag-of-birds approach to study robustness of automated audio based bird species classification.Appl.Sci.,2021,11(19):9226
• [3]Priyadarshani N,Marsland S,Castro I.Automated birdsong recognition in complex acoustic environments:a review.J.Avian Biol.,2018,49(5):jav-01447
• [4]Kasten E P,Gage S H,Fox J,et al.The remote environmental assessment laboratory’s acoustic library:an archive for studying soundscape ecology.Ecol.Inform.,2012,12:50-67
• [5]Go??au H,Glotin H,Vellinga W P,et al.LifeCLEF bird identification task2016:the arrival of deep learning.Conference and Labs of the Evaluation Forum 2016,CLEF Initiative,??vora,Portugal:2016:440-449
• [6]Hill S D,Ji W H,Parker K A,et al.A comparison of vocalisations between mainland tui (Prosthemadera novaeseelandiae novaeseelandiae) and Chatham Island tui (P.n.chathamensis).N.Z.J.Ecol.,2013,37(2):214-223
• [7]Liu D D,Xiao H G,Chen K.Research progress in bird sounds recognition based on acoustic monitoring technology:a systematic review.Appl.Acoust.,2025,228:110285
• [8]Cheng F,Chen J B,Qiu J F,et al.A subregion division based multi-objective evolutionary algorithm for SVM training set selection.Neurocomputing,2020,394:70-83
• [9]Choi H S,Kim J H,Lee E H,et al.Development of a revised multilayer perceptron model for dam inflow prediction.Water,2022,14(12):1878
• [10]Lee C H,Hsu S B,Shih J L,et al.Continuous birdsong recognition using Gaussian mixture modeling of image shape features.IEEE Trans.Multimed.,2013,15(2):454-464
• [11]Pérez-Granados C,Bota G,Giralt D,et al.A cost-effective protocol for monitoring birds using autonomous recording units:a case study with a night-time singing passerine.Bird Study,2018,65(3):338-345
• [12]Stowell D,PetruskováT,??álek M,et al.Automatic acoustic identification of individuals in multiple species:improving identification across recording conditions.J.R.Soc.Interface,2019,16(153):20180940
• [13]Mporas I,Ganchev T,Kocsis O,et al.Automated acoustic classification of bird species from real-field recordings.The 24th International Conference on Tools with Artificial Intelligence,IEEE,Athens,Greece,2012:778-781
• [14]Arriaga J G,Sanchez H,Vallejo E E,et al.Identification of Cassin’s Vireo (Vireo cassinii) individuals from their acoustic sequences using an ensemble of learners.Neurocomputing,2016,175:966-979
• [15]Xie J J,Zhong Y J,Zhang J G,et al.A review of automatic recognition technology for bird vocalizations in the deep learning era.Ecol.Inform.,2023,73:101927
• [16]Vellinga W P,PlanquéR.The Xeno-canto collection and its relation to sound recognition and classification.Conference and Labs of the Evaluation Forum 2015,CLEF Initiative,Toulouse,France:2015:1-8
• [17]Kahl S,Denton T,Klinck H,et al.Overview of BirdCLEF 2021:bird call identification in soundscape recordings.Conference and Labs of the Evaluation Forum 2021,CLEF Initiative,Bucharest,Romania,2021:1437-1450
• [18]Sullivan B L,Wood C L,Iliff M J,et al.eBird:a citizen-based bird observation network in the biological sciences.Biol.Conserv.,2009,142(10):2282-2292
• [19]Stowell D,Wood M D,Pamu??a H,et al.Automatic acoustic detection of birds through deep learning:the first bird audio detection challenge.Meth.Ecol.Evol.,2019,10(3):368-380
• [20]Zhang S K,Gao Y,Cai J M,et al.A novel bird sound recognition method based on multifeature fusion and a transformer encoder.Sensors,2023,23(19):8099
• [21]Solomes A,Stowell D.Efficient bird sound detection on the Bela embedded system.International Conference on Acoustics,Speech and Signal Processing (ICASSP),IEEE,Barcelona,Spain,2020:746-750
• [22]Xie J J,Hao Z L,Hu C H,et al.Beyond amplitude:phase integration in bird vocalization recognition with MHAResNet.Avian Res.,2025,16(1):100229
• [23]Pérez-Granados C,Schuchmann K L.Diel and annual patterns of vocal activity of three neotropical wetland birds revealed via BirdNET.Diversity,2025,17(5):324
• [24]Permana S D H,Saputra G,Arifitama B,et al.Classification of bird sounds as an early warning method of forest fires using convolutional neural network (CNN) algorithm.J.King Saud Univ.Comput.Inf.Sci.,2022,34(7):4345-4357
• [25]Wang K G,Wang L,Sun J D.The data analysis of sports training by ID3 decision tree algorithm and deep learning.Sci.Rep.,2025,15:15060
• [26]Wu Y C,Feng J W.Development and application of artificial neural network.Wirel.Pers.Commun.,2018,102(2):1645-1656
• [27]Sze V,Chen Y H,Yang T J,et al.Efficient processing of deep neural networks:a tutorial and survey.Proc.IEEE,2017,105(12):2295-2329
• [28]LeCun Y,Bottou L,Bengio Y,et al.Gradient-based learning applied to document recognition.Proc.IEEE,1998,86(11):2278-2324
• [29]Yu Y,Si X S,Hu C H,et al.A review of recurrent neural networks:LSTM cells and network architectures.Neural Comput.,2019,31(7):1235-1270
• [30]Van Houdt G,Mosquera C,Nápoles G.A review on the long shortterm memory model.Artif.Intell.Rev.,2020,53(8):5929-5955
• [31]Trevisan De Souza V L,Marques B A D,Batagelo H C,et al.A review on generative adversarial networks for image generation.Comput.Graph.,2023,114:13-25
• [32]Anidjar O H,Barak A,Ben-Moshe B,et al.A stethoscope for drones:transformers-based methods for UAVs acoustic anomaly detection.IEEE Access,2023,11:33336-33353
• [33]Hinton G E,Salakhutdinov R R.Reducing the dimensionality of data with neural networks.Science,2006,313(5786):504-507
• [34]Gers F A,Schmidhuber J,Cummins F.Learning to forget:continual prediction with LSTM.Neural Comput.,2000,12(10):2451-2471
• [35]Hochreiter S,Schmidhuber J.Long short-term memory.Neural Comput.,1997,9(8):1735-1780
• [36]Cho K,Van Merrienboer B,Gulcehre C,et al.Learning phrase representations using RNN encoder-decoder for statistical machine translation.Conference on Empirical Methods in Natural Language Processing (EMNLP),ACL,Doha,Qatar,2014:1724-1734
• [37]Zhang L E,Huang H N,Yin L,et al.Dolphin vocal sound generation via deep WaveGAN.J.Electron.Sci.Technol.,2022,20(3):100171
• [38]Bahmei B,Birmingham E,Arzanpour S.CNN-RNN and data augmentation using deep convolutional generative adversarial network for environmental sound classification.IEEE Signal Process.Lett.,2022,29:682-686
• [39]Fu Y X,Yu C J,Zhang Y,et al.Classification of birdsong spectrograms based on DR-ACGAN and dynamic convolution.Ecol.Inform.,2023,77:102250
• [40]Han K,Wang Y H,Chen H T,et al.A survey on vision transformer.IEEE Trans.Pattern Anal.Mach.Intell.,2023,45(1):87-110
• [41]Snyder Q,Jiang Q T,Tripp E.Integrating self-attention mechanisms in deep learning:a novel dual-head ensemble transformer with its application to bearing fault diagnosis.Signal Process.,2025,227:109683
• [42]Gong Y,Chung Y A,Glass J.AST:audio spectrogram transformer.Interspeech 2021,ISCA,Brno,Czech Republic,2021:571-575
• [43]Chen S Y,Wu Y,Wang C Y,et al.BEATs:audio pre-training with acoustic tokenizers.The 40th International Conference on Machine Learning (ICML),IMLS,Honolulu,Hawaii,USA,2023:5178-5193
• [44]Elizalde B,Deshmukh S,Al Ismail M,et al.CLAP learning audio concepts from natural language supervision.International Conference on Acoustics,Speech and Signal Processing (ICASSP),IEEE,Rhodes Island,Greece,2023:1-5
• [45]Gemmeke J F,Ellis D P W,Freedman D,et al.Audio Set:an ontology and human-labeled dataset for audio events.International Conference on Acoustics,Speech and Signal Processing(ICASSP),IEEE,New Orleans,LA,2017:776-780
• [46]Berwick R C,Okanoya K,Beckers G J L,et al.Songs to syntax:the linguistics of birdsong.Trends Cogn.Sci.,2011,15(3):113-121
• [47]Ghani B,Denton T,Kahl S,et al.Global birdsong embeddings enable superior transfer learning for bioacoustic classification.Sci.Rep.,2023,13:22876
• [48]Gavali P,Banu J S.A novel approach to Indian bird species identification:employing visual-acoustic fusion techniques for improved classification accuracy.Front.Artif.Intell.,2025,8:1527299
• [49]Ali H,Tran S N,Benetos E,et al.Speaker recognition with hybrid features from a deep belief network.Neural Comput.Appl.,2018,29(6):13-19
• [50]Knight E C,Hannah K C,Foley G J,et al.Recommendations for acoustic recognizer performance assessment with application to five common automated signal recognition programs.Avian Conserv.Ecol.,2017,12(2):14
• [51]Xie J,Zhu M Y.Handcrafted features and late fusion with deep learning for bird sound classification.Ecol.Inform.,2019,52:74-81
• [52]Wang D,Du B,Zhang L P,et al.Hyperspectral image classification based on multi-scale information compensation.Remote Sens.Lett.,2020,11(3):293-302
• [53]LeBien J,Zhong M,Campos-Cerqueira M,et al.A pipeline for identification of bird and frog species in tropical soundscape recordings using a convolutional neural network.Ecol.Inform.,2020,59:101113
• [54]Nanni L,Maguolo G,Paci M.Data augmentation approaches for improving animal audio classification.Ecol.Inform.,2020,57:101084
• [55]Zhong M,LeBien J,Campos-Cerqueira M,et al.Multispecies bioacoustic classification using transfer learning of deep convolutional neural networks with pseudo-labeling.Appl.Acoust.,2020,166:107375
• [56]Knight E C,Poo Hernandez S,Bayne E M,et al.Pre-processing spectrogram parameters improve the accuracy of bioacoustic classification using convolutional neural networks.Bioacoustics,2020,29(3):337-355
• [57]Maegawa Y,Ushigome Y,Suzuki M,et al.A new survey method using convolutional neural networks for automatic classification of bird calls.Ecol.Inform.,2021,61:101164
• [58]Zhong M,Taylor R,Bates N,et al.Acoustic detection of regionally rare bird species through deep convolutional neural networks.Ecol.Inform.,2021,64:101333
• [59]Nanni L,Maguolo G,Brahnam S,et al.An ensemble of convolutional neural networks for audio classification.Appl.Sci.,2021,11(13):5796
• [60]Bravo Sanchez F J,Hossain M R,English N B,et al.Bioacoustic classification of avian calls from raw sound waveforms with an open-source deep learning architecture.Sci.Rep.,2021,11:15733
• [61]Zhang F Y,Zhang L Y,Chen H X,et al.Bird species identification using spectrogram based on multi-channel fusion of DCNNs.Entropy,2021,23(11):1507
• [62]Dai Y S,Yang J,Dong Y W,et al.Blind source separation-based IVA-Xception model for bird sound recognition in complex acoustic environments.Electron.Lett.,2021,57(11):454-456
• [63]Gupta G,Kshirsagar M,Zhong M,et al.Comparing recurrent convolutional neural networks for large scale bird species classification.Sci.Rep.,2021,11(1):17085
• [64]Eichinski P,Alexander C,Roe P,et al.A convolutional neural network bird species recognizer built from little data by iteratively training,detecting,and labeling.Front.Ecol.Evol.,2022,10:810330
• [65]Kumar Y,Gupta S,Singh W.A novel deep transfer learning models for recognition of birds sounds in different environment.Soft Comput.,2022,26(3):1003-1023
• [66]Nolan V,Scott C,Yeiser J M,et al.The development of a convolutional neural network for the automatic detection of northern bobwhite Colinus virginianus covey calls.Remote Sens.Ecol.Conserv.,2022,9(1):46-61
• [67]Liu J,Zhang Y,Lv D,et al.Birdsong classification based on ensemble multi-scale convolutional neural network.Sci.Rep.,2022,12(1):8636
• [68]Lauha P,Somervuo P,Lehikoinen P,et al.Domain-specific neural networks improve automated bird sound recognition already with small amount of local data.Meth.Ecol.Evol.,2022,13(12):2799-2810
• [69]Xie S S,Lu J,Liu J,et al.Multi-view features fusion for birdsong classification.Ecol.Inform.,2022,72:101893
• [70]Xie J,Zhu M Y.Sliding-window based scale-frequency map for bird sound classification using 2D-and 3D-CNN.Expert Syst.Appl.,2022,207:118054
• [71]Hu T,Yuan M M,Li J H,et al.Avian vocalizations in Huangmaohai Sea-Crossing Channel:automatic birdsong recognition and ecological impact analysis based on deep learning.Biol.Conserv.,2025,305:111101
• [72]Van Osta J M,Dreis B,Meyer E,et al.An active learning framework and assessment of inter-annotator agreement facilitate automated recogniser development for vocalisations of a rare species,the southern black-throated finch (Poephila cincta cincta).Ecol.Inform.,2023,77:102233
• [73]Wolfe B,Proctor M D,Nolan V,et al.An efficient acoustic classifier for high-priority avian species in the southern Great Plains using convolutional neural networks.Wildl.Soc.Bull.,2023,47(4):e1492
• [74]MacLean K,Triguero I.Identifying bird species by their calls in Soundscapes.Appl.Intell.,2023,53(19):21485-21499
• [75]Jeantet L,Dufourq E.Improving deep learning acoustic classifiers with contextual information for wildlife monitoring.Ecol.Inform.,2023,77:102256
• [76]Noumida A,Rajan R.Multi-label bird species classification using sequential aggregation strategy from audio recordings.Comput.Inform.,2023,42(5):1255-1280
• [77]Qiu Z B,Wang H X,Liao C B,et al.Sound recognition of harmful bird species related to power grid faults based on VGGish transfer learning.J.Electr.Eng.Technol.,2023,18(3):2447-2456
• [78]Baowaly M K,Sarkar B C,Walid M A A,et al.Deep transfer learning-based bird species classification using Mel spectrogram images.PLoS One,2024,19(8):e0305708
• [79]Shen X H,Li G Y,Shi H F,et al.Ensemble learning strategy for birdsong recognition under data imbalance.Biodivers.Sci.,2024,32(10):24215
• [80]Kumar A S,Schlosser T,Kahl S,et al.Improving learning-based birdsong classification by utilizing combined audio augmentation strategies.Ecol.Inform.,2024,82:102699
• [81]Swaminathan B,Jagadeesh M,Vairavasundaram S.Multi-label classification for acoustic bird species detection using transfer learning approach.Ecol.Inform.,2024,80:102471
• [82]Xie S S,Xie J J,Zhang J G,et al.MDF-Net:a multi-view dualattention fusion network for efficient bird sound classification.Appl.Acoust.,2024,225:110138
• [83]Ghaffari H,Devos P.On the role of audio frontends in bird species recognition.Ecol.Inform.,2024,81:102573
• [84]Segura-Garcia J,Sturley S,Arevalillo-Herraez M,et al.5G AI-IoT system for bird species monitoring and song classification.Sensors,2024,24(11):3687
• [85]Wu W,Zhang R Y,Zheng X Y,et al.Orchard bird song recognition based on multi-view multi-level contrastive learning.Appl.Acoust.,2024,224:110133
• [86]Su S,Gu D H,Lai J Y,et al.A novel deep learning-based bioacoustic approach for identification of look-alike white-eye (Zosterops) species traded in wildlife markets.Ibis,2025,167(1):41-55
• [87]Heinrich R,Rauch L,Sick B,et al.AudioProtoPNet:an interpretable deep learning model for bird sound classification.Ecol.Inform.,2025,87:103081
• [88]Kazeneza M,Bosman A S,Amenyedzi D K,et al.Balancing complexity and performance of machine learning models for avian pests sound detection in agricultural environments.IEEE Access,2025,13:105813-105827
• [89]Ghani B,Kalkman V J,PlanquéB,et al.Impact of transfer learning methods and dataset characteristics on generalization in birdsong classification.Sci.Rep.,2025,15(1):16273
• [90]Han X,Peng J X.Multi-label bird species classification using transfer learning network.Arch.Acoust.,2025,50(2):223-233
• [91]Wu T H,Ma W,Gu O,et al.SeqFusionNet:a hybrid model for sequence-aware and globally integrated acoustic representation.PLoS One,2025,20(9):e0330691
• [92]León Du’Mottuchi X,Creanza N.The role of learned song in the evolution and speciation of Eastern and Spotted towhees.PLoS Comput.Biol.,2025,21(6):e1013135
• [93]Wang Y N,Chen A B,Li H C,et al.A hierarchical birdsong feature extraction architecture combining static and dynamic modeling.Ecol.Indic.,2023,150:110258
• [94]Zhang X,Chen A B,Zhou G X,et al.Spectrogram-frame linear network and continuous frame sequence for bird sound classification.Ecol.Inform.,2019,54:101009
• [95]Mohanty R,Kumar Mallik B,Singh Solanki S.Recognition of bird species based on spike model using bird dataset.Data Brief,2020,29:105301
• [96]Rajan R,Johnson J,Abdul Kareem N.Bird call classification using DNN-based acoustic modelling.Circuits Syst.Signal Process.,2022,41(5):2669-2680
• [97]Wang H L,Xu Y F,Yu Y,et al.An efficient model for a vast number of bird species identification based on acoustic features.Animals,2022,12(18):2434
• [98]Kahl S,Wood C M,Eibl M,et al.BirdNET:a deep learning solution for avian diversity monitoring.Ecol.Inform.,2021,61:101236
• [99]Chan T K,Chin C S.A comprehensive review of polyphonic sound event detection.IEEE Access,2020,8:103339-103373
• [100]Dufourq E,Durbach I,Hansford J P,et al.Automated detection of Hainan gibbon calls for passive acoustic monitoring.Remote Sens.Ecol.Conserv.,2021,7(3):475-487
• [101]Funosas D,Barbaro L,SchilléL,et al.Assessing the potential of BirdNET to infer European bird communities from large-scale ecoacoustic data.Ecol.Indic.,2024,164:112146
• [102]Stowell D.Computational bioacoustics with deep learning:a review and roadmap.PeerJ,2022,10:e13152
• [103]Ruff Z J,Lesmeister D B,Duchac L S,et al.Automated identification of avian vocalizations with deep convolutional neural networks.Remote Sens.Ecol.Conserv.,2019,6(1):79-92
• [104]Florentin J,Dutoit T,Verlinden O.Detection and identification of European woodpeckers with deep convolutional neural networks.Ecol.Inform.,2020,55:101023
• [105]Morgan M M,Braasch J.Long-term deep learning-facilitated environmental acoustic monitoring in the Capital Region of New York State.Ecol.Inform.,2021,61:101242
• [106]Gillings S,Scott C.Nocturnal flight calling behaviour of thrushes in relation to artificial light at night.Ibis,2021,163(4):1379-1393
• [107]Morales G,Vargas V,Espejo D,et al.Method for passive acoustic monitoring of bird communities using UMAP and a deep neural network.Ecol.Inform.,2022,72:101909
• [108]Cole J S,Michel N L,Emerson S A,et al.Automated bird sound classifications of long-duration recordings produce occupancy model outputs similar to manually annotated data.Ornithol.Appl.,2022,124(2):duac003
• [109]Hane M E,Thornton-Frost J E,Springford A,et al.Factors associated with automated detection of northern spotted owl (Strix occidentalis caurina) four-note location calls.Avian Conserv.Ecol.,2022,17:art26
• [110]Folliot A,Haupert S,Ducrettet M,et al.Using acoustics and artificial intelligence to monitor pollination by insects and tree use by woodpeckers.Sci.Total Environ.,2022,838:155883
• [111]Bota G,Manzano-Rubio R,Catalán L,et al.Hearing to the unseen:AudioMoth and BirdNET as a cheap and easy method for monitoring cryptic bird species.Sensors,2023,23(16):7176
• [112]Van Doren B M,Farnsworth A,Stone K,et al.Nighthawk:acoustic monitoring of nocturnal bird migration in the Americas.Meth.Ecol.Evol.,2023,15(2):329-344
• [113]Clark M L,Salas L,Baligar S,et al.The effect of soundscape composition on bird vocalization classification in a citizen science biodiversity monitoring project.Ecol.Inform.,2023,75:102065
• [114]Allen-Ankins S,Mula Laguna J,Schwarzkopf L.Towards automated detection of the endangered southern black-throated finch(Poephila cincta cincta).Wildl.Res.,2024,51(10):WR23151
• [115]Duarte A,Weldy M J,Lesmeister D B,et al.Passive acoustic monitoring and convolutional neural networks facilitate high-resolution and broadscale monitoring of a threatened species.Ecol.Indic.,2024,162:112016
• [116]Lavner Y,Melamed R,Bashan M,et al.The bioacoustic soundscape of a pandemic in Agmon Hula lake park:continuous annual monitoring using a deep learning system.Ecol.Inform.,2024,80:102528
• [117]Poutaraud J,Sueur J,Thébaud C,et al.Meta-Embedded Clustering(MEC):a new method for improving clustering quality in unlabeled bird sound datasets.Ecol.Inform.,2024,82:102687
• [118]Lostanlen V,Cramer A,Salamon J,et al.BirdVoxDetect:largescale detection and classification of flight calls for bird migration monitoring.IEEE/ACM Trans.Audio Speech Lang.Process.,2024,32:4134-4145
• [119]Zhang C Y,Jin N T,Xie J,et al.CicadaNet:deep learning based automatic cicada chorus filtering for improved long-term bird monitoring.Ecol.Indic.,2024,158:111423
• [120]Fradet D T,Cimino M A,White E R,et al.Designing a BirdNET classifier for high wind detection in passive acoustic recordings to support wildlife monitoring.J.Acoust.Soc.Am.,2025,157(6):4502-4512
• [121]Huus J,Kelly K G,Bayne E M,et al.HawkEars:a regional,highperformance avian acoustic classifier.Ecol.Inform.,2025,87:103122
• [122]Moreno