Combining interactive AI and ecological expertise to scale biodiversity observation across ecosystems.

Our research in wildlife monitoring explores how interactive and adaptive artificial intelligence can transform biodiversity observation at scale. We develop machine learning systems that work with experts—rather than replacing them—to accelerate the analysis of large, unlabeled environmental datasets such as Passive Acoustic Monitoring (PAM) recordings and other sensor streams.

Through collaborations with ecologists, conservation agencies, and biosphere reserve networks, we design tools that integrate human-in-the-loop learning, explainable AI, and visual analytics. These systems help identify patterns in vast soundscapes, enabling early detection of species, tracking of ecosystem health, and informed conservation decisions.

Our work has been recognized internationally, including a Top-3 placement in the $10 M XPRIZE Rainforest competition, and continues through projects such as Chorus RF, YAPAT, and EcoScape Analyzer, as well as partnerships with institutions like ICMBio for long-term biodiversity monitoring in UNESCO Biosphere Reserves.

Contact

Thiago Gouvêa, thiago.gouvea@dfki.de