In a forest near Nairobi, the capital of Kenya, researchers at Kenya's Dedan Kimati University of Science and Technology are debugging bioacoustic equipment equipped with highly sensitive microphones that can clearly record bird calls.
This recording equipment is powered by solar energy and works between 5 a.m. and 10 a.m. and between 17 p.m. and 19 p.m. every day.
Jason Nyorog, a junior researcher at the Data Science and Artificial Intelligence Center at Dedan Kimati University of Science and Technology in Kenya, said that these recorded bird "songs" will be converted into spectrograms and provided to the artificial intelligence model. The model can determine which birds are in the recorded audio files.
Researchers say the presence or absence of certain indicator species can provide clues to the health of forest ecosystems.
The Kenya Bird Map project, launched by the National Museum of Kenya, aims to map the distribution of birds in Kenya.
Samuel, the leader of Kenya's "Bird Map" project, said that by analyzing changes in bird species and numbers, it can help scientists understand the situation of the habitat and further understand how changes in the habitat environment affect the survival of birds.
Samuel: "When a place is not suitable for birds to breed or feed, or even to occupy a territory, they will fly away. Birds can usually do this faster than most animals because they have flight capabilities, and their wide range of movement, so the presence or absence of specific species in a forest ecosystem can be an important indicator of the health of that forest ecosystem."
According to Samuel, given current trends in human activities and development, some vulnerable bird species may face a 20%-30% reduction in habitat over the next decade, while those that are endangered may Facing extinction. Therefore, it is of great significance to help identify changes in bird populations through artificial intelligence technology.
