# Goal of the competition Mobile and habitat-diverse species serve as valuable indicators of biodiversity change, as shifts in their assemblages and population dynamics can signal the success or failure of ecological restoration efforts. However, conducting traditional observer-based biodiversity surveys across large areas is both costly and logistically demanding. In contrast, passive acoustic monitoring (PAM), combined with modern machine learning techniques, enables conservationists to sample across broader spatial scales with greater temporal resolution, providing deeper insights into the relationship between restoration interventions and biodiversity. For this competition, you'll apply your machine-learning expertise to identify under-studied species based on their acoustic signatures. Specifically, you'll develop computational methods to process continuous audio data and recognize species from different taxonomic groups by their sounds. The most effective solutions will demonstrate the ability to train reliable classifiers with limited labeled data. If successful, your work will contribute to ongoing efforts to enhance biodiversity monitoring, including research initiatives in the lowlands of the Magdalena Valley of Colombia. # Context Humid tropical rainforests, Earth's most biodiverse and ancient ecosystems, are vital for climate regulation and water resource protection. However, rainforests face severe threats. In Colombia, a megadiverse country, the lowlands of the Magdalena Valley are a biodiversity hotspot and home to many endangered species. Over 70% of the Magdalena Valley lowland rainforests are replaced by vast pastures for cattle ranching, and illegal logging is common in forest fragment remnants. The protection of the last forest remnants and wetlands is an urgent need. Fundación Biodiversa Colombia (FBC) collaborates with local communities, landowners, and organizations to conserve, restore, and connect fragments of forests and wetlands. Established in 2012, [El Silencio Natural Reserve](https://www.fundacionbiodiversa.org/fundacion2024/information/) protects 5,407 acres of tropical lowland forests and wetlands. Home to diverse wildlife, including 295 birds, 34 amphibians, 69 mammals, 50 reptiles, and nearly 500 plant species, El Silencio is a model for regional conservation and sustainability. A significant part of the reserve, previously used for extensive livestock farming, is under an ecological restoration project. Through the Kaggle competition, we aim to automate detecting and classifying different taxonomic groups of soundscapes from El Silencio Natural Reserve, intending to provide a better understanding of the ecological process of the restoration projects. The broader goals for this Kaggle competition include: (1) Identify species of different taxonomic groups in the Middle Magdalena Valley of Colombia/El Silencio Natural Reserve in soundscape data. (2) Train machine learning models with very limited amounts of training samples for rare and endangered species. (3) Enhance machine learning models with unlabeled data for improving detection/classification. Thanks to your innovations, it will be easier for researchers and conservation practitioners to understand restoration activities' effect trends accurately. As a result, they'll be able to evaluate threats and adjust their conservation actions regularly and more effectively. This competition is collaboratively organized by (alphabetic order) the Chemnitz University of Technology, Fundación Biodiversa Colombia, Google Research, iNaturalist, Instituto Humboldt, K. Lisa Yang Center for Conservation Bioacoustics at the Cornell Lab of Ornithology, LifeCLEF, Red Ecoacústica Colombiana, University College London, and Xeno-canto.