Burooj Ghani

Biodiversity is declining at an alarming rate: the WWF Living Planet Index reports ~69% decline in monitored vertebrate populations since 1970, and ~1 million species are threatened with extinction (IPBES). My work is motivated by the need to build tools that provide reliable estimates of ecosystem health. I focus on sound because it scales: it captures soniferous life continuously, non-invasively, and across vast areas. My research combines machine learning with ecology and bioacoustics to make sound a core tool for biodiversity monitoring.

I trained as a computer scientist, drawn to algorithms and modeling problems. Later, I obtained my PhD in Computer Science from the University of Göttingen, where I ventured into bioacoustics. Since then I have specialized in computational bioacoustics, from advancing general-purpose representations for bioacoustics to developing open-source models such as AvesEcho. The model was integrated into Xeno-canto, the world’s biggest bird sound forum. To me, that was an important moment - it was no longer just an academic model, but something that birders and conservationists could actually use. More recently, I co-launched InsectSet459, an insect sound deep learning benchmark, to broaden acoustic monitoring beyond birds. My broader vision is to build computational tools that help us better understand and protect the living world.