A list of useful resources in the bird sound recognition – bird songs & calls

• Datasets
• Papers
• Open Source Projects
• Competitions
• Articles

Feel free to make a pull request or to  the repository if you like it!

Introduction

What are challenges in bird song recognition? Elias Sprengel, Martin Jaggi, Yannic Kilcher, and Thomas Hofmann in their paper Audio Based Bird Species Identification using Deep Learning Techniques point out some very important issues:

• Background noise in the recordings – city noises, churches, cars…
• Very often multiple birds singing at the same time – multi-label classification problem
• Differences between mating calls and songs – mating calls are short, whereas songs are longer
• Inter-species variance – same bird species singing in different countries might sound completely different
• Variable length of sound recordings
• Large number of different species

Datasets

• xeno-canto.org is a website dedicated to sharing bird sounds from all over the world (480k, September 2019). Scripts that make downloading easier can be found here:

  • github.com/AgaMiko/xeno-canto-download – Simple and easy scraper to download sound with metadata, written in python
  • github.com/ntivirikin/xeno-canto-py – Python API wrapper designed to help users easily download xeno-canto.org recordings and associated information. Avaiable to install with pip manager.

• Macaulay Library is the world’s largest archive of animal sounds. It includes more than 175,000 audio recordings covering 75 percent of the world’s bird species. There are an ever-increasing numbers of insect, fish, frog, and mammal recordings. The video archive includes over 50,000 clips, representing over 3,500 species.[1] The Library is part of Cornell Lab of Ornithology of the Cornell University.

• tierstimmenarchiv.de – Animal sound album at the Museum für Naturkunde in Berlin, with a collection of bird songs and calls.

• RMBL-Robin database – Database for Noise Robust Bird Song Classification, Recognition, and Detection.A 78 minutes Robin song database collected by using a close-field song meter (www.wildlifeacoustics.com) at the Rocky Mountain Biological Laboratory near Crested Butte, Colorado in the summer of 2009. The recorded Robin songs are naturally corrupted by different kinds of background noises, such as wind, water and other vocal bird species. Non-target songs may overlap with target songs. Each song usually consists of 2-10 syllables. The timing boundaries and noise conditions of the syllables and songs, and human inferred syllable patterns are annotated.

• floridamuseum.ufl.edu/bird-sounds – A collection of bird sound recordings from the Florida Museum Bioacoustic Archives, with 27,500 cataloged recordings representing about 3,000 species, is perhaps third or fourth largest in the world in number of species.

• Field recordings, worldwide (“freefield1010”) – a collection of 7,690 excerpts from field recordings around the world, gathered by the FreeSound project, and then standardised for research. This collection is very diverse in location and environment, and for the BAD Challenge we have annotated it for the presence/absence of birds.

  • Download: data labels • audio files (5.8 Gb zip) (or via bittorrent)

• Crowdsourced dataset, UK (“warblrb10k”) – 8,000 smartphone audio recordings from around the UK, crowdsourced by users of Warblr the bird recognition app. The audio covers a wide distribution of UK locations and environments, and includes weather noise, traffic noise, human speech and even human bird imitations.

  • Download: data labels • audio files (4.3 Gb zip) (or via bittorrent)

• Remote monitoring flight calls, USA (“BirdVox-DCASE-20k”) – 20,000 audio clips collected from remote monitoring units placed near Ithaca, NY, USA during the autumn of 2015, by the BirdVox project. More info about BirdVox-DCASE-20k

  • Download: data labels • audio files (15.4 Gb zip)

• british-birdsongs.uk – A collection of bird songs, calls and alarms calls from Great Britain

• birding2asia.com/W2W/freeBirdSounds – Bird recordigns from India, Philippines, Taiwan and Thailad.

• azfo.org/SoundLibrary/sounds_library – All recordings are copyrighted© by the recordist. Downloading and copying are authorized for noncommercial educational or personal use only.

Feel free to add other datasets to a list if you know any!

Papers

2020

• Priyadarshani, Nirosha, et al. “Wavelet filters for automated recognition of birdsong in long‐time field recordings.” Methods in Ecology and Evolution 11.3 (2020): 403-417.     
  
  Abstract

• Brooker, Stuart A., et al. “Automated detection and classification of birdsong: An ensemble approach.” Ecological Indicators 117 (2020): 106609.     
  
  Abstract

2019

• Stowell, Dan, et al. “Automatic acoustic detection of birds through deep learning: the first Bird Audio Detection challenge.” Methods in Ecology and Evolution 10.3 (2019): 368-380.     
  
  Abstract

• Koh, Chih-Yuan, et al. “Bird Sound Classification using Convolutional Neural Networks.” (2019).     
  
  Abstract

• Kahl, S., et al. “Overview of BirdCLEF 2019: large-scale bird recognition in Soundscapes.” CLEF working notes (2019).     
  
  Abstract

2018

• Kojima, Ryosuke, et al. “HARK-Bird-Box: A Portable Real-time Bird Song Scene Analysis System.” 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2018.     
  
  Abstract

• Fazeka, Botond, et al. “A multi-modal deep neural network approach to bird-song identification.” arXiv preprint arXiv:1811.04448 (2018).     
  
  Abstract

• Lasseck, Mario. “Audio-based Bird Species Identification with Deep Convolutional Neural Networks.” CLEF (Working Notes). 2018.     
  
  Abstract

• Priyadarshani, Nirosha, Stephen Marsland, and Isabel Castro. “Automated birdsong recognition in complex acoustic environments: a review.” Journal of Avian Biology 49.5 (2018): jav-01447.     
  
  Abstract

• Goeau, Herve, et al. “Overview of BirdCLEF 2018: monospecies vs. soundscape bird identification.” 2018.     
  
  Abstract

2017

• Zhao, Zhao, et al. “Automated bird acoustic event detection and robust species classification.” Ecological Informatics 39 (2017): 99-108.     
  
  Abstract

• Hershey, S. et. al., CNN Architectures for Large-Scale Audio Classification, ICASSP 2017

• Gemmeke, J. et. al., AudioSet: An ontology and human-labelled dataset for audio events, ICASSP 2017

• Salamon, Justin, et al. “Fusing shallow and deep learning for bioacoustic bird species classification.” 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2017.

  Abstract

• Narasimhan, Revathy, Xiaoli Z. Fern, and Raviv Raich. “Simultaneous segmentation and classification of bird song using CNN.” 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2017.

  Abstract

• Grill, Thomas, and Jan Schlüter. “Two convolutional neural networks for bird detection in audio signals.” 2017 25th European Signal Processing Conference (EUSIPCO). IEEE, 2017.

  Abstract

2016

• Tóth, Bálint Pál, and Bálint Czeba. Convolutional Neural Networks for Large-Scale Bird Song Classification in Noisy Environment, CLEF (Working Notes). 2016.

  Abstract

• Nicholson, David. “Comparison of machine learning methods applied to birdsong element classification.” Proceedings of the 15th Python in Science Conference. 2016.

  Abstract

• Sprengel, Elias, et al. Audio based bird species identification using deep learning techniques. No. CONF. 2016.

  Abstract

• Stowell, Dan, et al. “Bird detection in audio: a survey and a challenge.” 2016 IEEE 26th International Workshop on Machine Learning for Signal Processing (MLSP). IEEE, 2016.     
  
  Abstract

2015

• Tan, Lee N., et al. “Dynamic time warping and sparse representation classification for birdsong phrase classification using limited training data.” The Journal of the Acoustical Society of America 137.3 (2015): 1069-1080.     
  
  Abstract

Competitions

• BirdCLEF 2021 – Birdcall Identification – identify which birds are calling in long recordings, given training data generated in meaningfully different contexts. This is the exact problem facing scientists trying to automate the remote monitoring of bird populations. This competition builds on the previous one by adding soundscapes from new locations, more bird species, richer metadata about the test set recordings, and soundscapes to the train set.
• kaggle – Cornell Birdcall Identification – Build tools for bird population monitoring. Identify a wide variety of bird vocalizations in soundscape recordings. Due to the complexity of the recordings, they contain weak labels. There might be anthropogenic sounds (e.g., airplane overflights) or other bird and non-bird (e.g., chipmunk) calls in the background, with a particular labeled bird species in the foreground. Bring your new ideas to build effective detectors and classifiers for analyzing complex soundscape recordings!
• LifeCLEF 2020 – BirdCLEF – Two scenarios will be evaluated: (i) the recognition of all specimens singing in a long sequence (up to one hour) of raw soundscapes that can contain tens of birds singing simultaneously, and (ii) chorus source separation in complex soundscapes that were recorded in stereo at very high sampling rate (250 kHz SR). The training set used for the challenge will be a version of the 2019 training set enriched by new contributions from the Xeno-canto network and a geographic extension. It will contain approximately 80K recordings covering between 1500 and 2000 species from North, Central and South America, as well as Europe. This will be the largest bioacoustic dataset used in the literature.
• LifeCLEF 2019 Bird Recognition – The goal of the challenge is to detect and classify all audible bird vocalizations within the provided soundscape recordings. Each soundscape is divided into segments of 5 seconds. Participants should submit a list of species associated with probability scores for each segment.
• LifeCLEF 2018 Bird – Monophone – The goal of the task is to identify the species of the most audible bird (i.e. the one that was intended to be recorded) in each of the provided test recordings. Therefore, the evaluated systems have to return a ranked list of possible species for each of the 12,347 test recordings.
• LifeCLEF 2018 Bird – Soundscape – The goal of the task is to localize and identify all audible birds within the provided soundscape recordings. Each soundscape is divided into segments of 5 seconds, and a list of species associated to probability scores will have to be returned for each segment.
• Bird audio detection DCASE2018 – The task is to design a system that, given a short audio recording, returns a binary decision for the presence/absence of bird sound (bird sound of any kind). The output can be just “0” or “1”, but we encourage weighted/probability outputs in the continuous range [0,1] for the purposes of evaluation. For the main assessment we will use the well-known “Area Under the ROC Curve” (AUC) measure of classification performance.
• Bird Audio Detection Challenge 2016–2017 – contest organized in collaboration with the IEEE Signal Processing Society. They propose a research data challenge to create a robust and scalable bird detection algorithm. Organizers offer new datasets collected in real live bioacoustics monitoring projects, and an objective, standardised evaluation framework – and prizes for the strongest submissions. Results and summary of the best submissions can be found here.

Open Source Projects

• Polish bird species recognition – 19 class recognition, 2019 – Repositorium with code to download, and cut files into melspectrograms with librosa library. Later, files are classified with deep neural networks in Keras.
• Large-Scale Bird Sound Classification using Convolutional Neural Networks, 2017 – Code repo for our submission to the LifeCLEF bird identification task BirdCLEF2017.
• Automatic recognition of element classes and boundaries in the birdsong with variable sequences – This is a source code for the manuscript “Automatic recognition of element classes and boundaries in the birdsong with variable sequences” by Takuya Koumura and Kazuo Okanoya (http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0159188).
• Trainig scripts for deep convolutional neural network based audio classification in Keras – The following scripts were created for the BirdCLEF 2016 competition by Bálint Czeba and Bálint Pál Tóth.
• BirdSong Recognition – Classification system based on the classic HMM+MFCC method. It has 22 kinds of birds now, the correct rate is 81.8% (with only 1.7G data trained).
• A model for bird sound classification – pretrained VGGish/ Audioset model by Google and finetune it by letting it iterate during training on more than 80,000 audio samples of 10 second length (195 bird classes)
• Bird brain – This repo contains code to search the Xeno-canto bird sound database, and train a machine learning model to classify birds according to those sounds.
• Bird-Species-Classification – The project uses a neural-net in tensorflow to classify the species to which a bird belongs to based on the features it has. There are total 312 features and 11787 examples.
• Bird Species Classification by song – This repository is not actively maintained. It is the result of a master’s thesis and the code has been made available as a reference if anyone would like to reproduce the results of the thesis.
• Recognizing Birds from Sound – The 2018 BirdCLEF Baseline System – a baseline system for the LifeCLEF bird identification task BirdCLEF2018. Authors encourage participants to build upon the code base and share their results for future reference. They promise to keep the repository updated and add improvements and submission boilerplate in the future.

Articles

• SOUND-BASED BIRD CLASSIFICATION – How a group of Polish women used deep learning, acoustics and ornithology to classify birds