The University of Puerto Rico - Rio Piedras is awarded a grant to develop an automated remote biodiversity monitoring network that remotely censuses bird and amphibians using sound. The project will produce machine-learning algorithms for species identification, a database, and a web accessible interface for dynamic content. The network will be deployed in tropical montane cloud forests in Puerto Rico where hourly recordings will be transferred wirelessly to remote analysis centers to be digitized. Species recognition programs will use turbo-coding and Hidden Markov Models. An open platform based on Linux, Apache, MySQL and PHP will provide the web dissemination of audio, identifications, and associated climate data collected simultaneously. The project will work closely with the Caribbean National Forest and the Puerto Rico Department of Natural Resources and Environment to ensure that the system is useful for them and other groups that may adopt it in the future. In addition, public outreach will be initiated by incorporating members of the Ornithological Society of Puerto Rico in the bird vocalization identification component of the project.
