Identify Bird Species in the Western Ghats is a specific application of audio classification, where the goal is to classify bird species based on their calls captured in audio recordings. Deep learning techniques, such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs), can be implemented to effectively classify bird species from audio data. Therefore, the project involves implementing deep learning models for audio classification to achieve the objective of identifying bird species in the Western Ghats region.
\n\nIntroduces the concept of audio and sound classification using machine learning and deep learning techniques. It highlights the significance of natural language processing (NLP) in enabling machines to understand human language, whether in the form of speech or text.
\n\nThis concept aligns well with the project \"Identify Bird Species in the Western Ghats,\" as it involves utilizing machine learning and potentially deep learning methods to classify bird species based on audio recordings, which are essentially representations of sound. By applying NLP principles, machines can analyze and interpret the audio data, extracting meaningful features to classify different bird species based on their distinct calls.
\n","metadata":{}},{"cell_type":"markdown","source":"The competition aims to address the challenges of monitoring biodiversity, particularly bird species, in the Western Ghats region using advanced technology and machine learning. By leveraging passive acoustic monitoring (PAM) and machine learning techniques, participants are tasked with automating the detection and classification of bird species based on soundscapes.
\n\nThe primary goal is to develop computational solutions capable of identifying under-studied Indian bird species by their calls. Participants are expected to process continuous audio data and train reliable classifiers with limited training data. Successful solutions will contribute to ongoing efforts to protect avian biodiversity in the Western Ghats, India.
\n\nThe Western Ghats, a Global Biodiversity Hotspot, is facing significant anthropogenic pressures such as habitat modification and climate change. This necessitates the use of advanced conservation tools and technologies to monitor biodiversity effectively. The competition aims to automate the detection and classification of bird species in this region, given its diverse ecosystems and high levels of bird diversity.
\n\nThe competition timeline spans from April 3, 2024, to June 10, 2024, with various deadlines for entry, team merger, and final submissions. All deadlines are at 11:59 PM UTC on the corresponding day unless otherwise noted.
\n\nEvaluation is based on a version of macro-averaged ROC-AUC, with classes that have no true positive labels skipped. Participants are required to submit predictions for each bird species present in the audio recordings, with one column per species.
\n\nParticipants have the option to submit working notes, which will be evaluated based on criteria such as originality, quality, contribution, and presentation. Each working note will be reviewed by two reviewers, and scores will be averaged.
\n\nThis structured overview provides key insights into the competition's objectives, context, timeline, evaluation criteria, and optional working note submission process.
Participants in the competition are invited to submit working notes to the LifeCLEF 2024 conference. A special BirdCLEF working note competition will be conducted during the conference, where participants can showcase their findings, methodologies, and insights gained from the competition.
\nThe top two best working note award winners will each receive a cash prize of $2,500.
\nParticipants should submit their working notes in accordance with the guidelines provided by the conference organizers.
\nThe evaluation of working notes will be based on various criteria, including originality, quality, contribution, and presentation. For detailed judging criteria, refer to the Evaluation page of the competition.
\nSubmissions to this competition must be made through Notebooks. For the \"Submit\" button to be active after a commit, the following conditions must be met:
\nPlease see the Code Competition FAQ for more information on how to submit. And review the code debugging doc if you encounter submission errors.
","metadata":{}},{"cell_type":"markdown","source":"train_audio/: The training data consists of short recordings of individual bird calls generously uploaded by users of xenocanto.org. These files have been downsampled to 32 kHz where applicable to match the test set audio and converted to the ogg format. The training data should have nearly all relevant files; we expect there is no benefit to looking for more on xenocanto.org and appreciate your cooperation in limiting the burden on their servers.
\ntest_soundscapes/: When you submit a notebook, the test_soundscapes directory will be populated with approximately 1,100 recordings to be used for scoring. They are 4 minutes long and in ogg audio format. The file names are randomized but have the general form of soundscape_xxxxxx.ogg. It should take your submission notebook approximately five minutes to load all of the test soundscapes.
\nunlabeled_soundscapes/: Unlabeled audio data from the same recording locations as the test soundscapes.
\ntrain_metadata.csv: A wide range of metadata is provided for the training data. The most directly relevant fields are:\n
sample_submission.csv: A valid sample submission.\n
eBird_Taxonomy_v2021.csv: Data on the relationships between different species.
\n