In Deep Voice, we believe data-driven research holds the key for conservation of endangered species. Specifically, utilizing data-driven algorithms for marine mammals’ communication study is of significant conservation value. Nevertheless, a substantial gap between data-based cutting edge-technologies and conservation exists. In Deep Voice, we seek to mitigate this gap for the sake of marine mammals’ welfare.
What we do?
To achieve that, we develop AI-based passive acoustics monitoring (PAM) framework for using conservation research groups worldwide and provide pro-bono services for its implementation. So far, we have developed marine mammals’ vocalizations detection and classification algorithms for PAM purposes. These algorithms currently assist conservation groups in achieving their research goals (see our projects) in Costa Rica, Mozambique, Columbia, Brazil, Belize, Canada, and the USA.
Currently, we’re looking forward to participating in and assisting new conservation projects and collaborators using our services ( see “Are you A biologist?” page) and expanding our map of impact (Add in the website)
How we do it?
As much as we value the correct algorithm, conservation is eventually made by the right people. We follow an ongoing communication protocol with our collaborators in an end-to-end manner. First, we better understand the research groups data, motivation, and local needs of conservation. Then, using our expertise (see who we are?), we specifically tailor the right algorithmic approach to the conservation problem at hand and provide data-based solutions.
To support the cause, our framework is accessible and easy to use by a designated website. Furthermore, supporting one-to-one service is provided by our volunteers, following our communication protocol. To enhance the impact of our work, we share our code freely as an open source for the sake of academic research.
Near Future Plans
Scale our algorithms usage and expand our map of impact. Contribute to at least five more conservation groups with detection\ classification services. Make localization and data-analysis algorithms accessible to biologists worldwide.
Assist in the development of marine mammals anti-collision technology. Establish an online community of researchers that use AI-based solutions for conservation. Find new strategic partnerships.
Five years from now
Deep Voice map of impact spreads all over the globe.
A worldwide network of bio-acousticians that follow a data-based paradigm of conservation is founded. An end-to-end density estimation framework is provided by Deep Voice
to research groups worldwide. This framework supports data-based regulations monitoring marine mammals welfare worldwide. A clear measure of marine mammals’ conservation success exists and is quantified.