BirdCLEF 2021 Kaggle Challenge

Join the BirdCLEF 2021 Kaggle challenge, where you’ll automate the acoustic identification of birds in soundscape recordings. You'll examine an acoustic dataset to build detectors and classifiers to extract the signals of interest (bird calls) by developing innovative solutions that are able to do so efficiently and reliably.

Enter by the team merger deadline, May 24, 2021.

Find the full challenge details, requirements, and entry information here.

Date published: 2021/04/19

Birds of a feather flock together. Thankfully, this makes it easier to hear them! There are over 10,000 bird species around the world. Identifying the red-winged blackbirds or Bewick’s wrens in an area, for example, can provide important information about the habitat. As birds are high up in the food chain, they are excellent indicators of deteriorating environmental quality and pollution. Monitoring the status and trends of biodiversity in ecosystems is no small task. With proper sound detection and classification—aided by machine learning—researchers can improve their ability to track the status and trends of biodiversity in important ecosystems, enabling them to better support global conservation efforts.

Recent advances in machine listening have improved acoustic data collection. However, it remains a challenge to generate analysis outputs with high precision and recall. The majority of data is unexamined due to a lack of effective tools for efficient and reliable extraction of the signals of interests (e.g., bird calls).

The Cornell Lab of Ornithology is dedicated to advancing the understanding and protection of birds and the natural world. The Lab joins with people from all walks of life to make new scientific discoveries, share insights, and galvanize conservation action. For this competition, they're collaborating with Google Research, LifeCLEF, and Xeno-canto.

In this competition, you’ll automate the acoustic identification of birds in soundscape recordings. You'll examine an acoustic dataset to build detectors and classifiers to extract the signals of interest (bird calls). Innovative solutions will be able to do so efficiently and reliably.

The ornithology community is collecting many petabytes of acoustic data every year, but the majority of data remains unexamined. If successful, you'll help researchers properly detect and classify bird sounds, significantly improving their ability to monitor the status and trends of biodiversity in important ecosystems. Researchers will better be able to infer factors about an area’s quality of life based on a changing bird population, which allows them to identify how they can best support global conservation efforts.

The LifeCLEF Bird Recognition Challenge (BirdCLEF) focuses on developing machine learning algorithms to identify avian vocalizations in continuous soundscape data to aid conservation efforts worldwide. Launched in 2014, it has become one of the largest bird sound recognition competitions in terms of dataset size and species diversity.

Enter the Challenge!

Enter by the team merger deadline, May 24, 2021.

Find the full challenge details, requirements, and entry information here.

Discuss bird acoustics in our forum!

Continue the discussion… Acoustic Monitoring