Artificial intelligence (AI) has rapidly become a transformative force in marine biology research, revolutionizing how scientists collect, analyze, and interpret data from the ocean. By processing vast, complex datasets, AI and its subfield, machine learning, are overcoming long-standing challenges related to the ocean's scale, complexity, and inaccessibility. This has accelerated discoveries, enhanced conservation efforts, and enabled more accurate predictions about the future of marine ecosystems.
Data analysis and species identification
A primary influence of AI has been its ability to automate the analysis of visual and acoustic data, a task that was once time-consuming and labor-intensive for human researchers.
Automated image processing: Algorithms can be trained on extensive datasets, such as the open-source FathomNet, to automatically detect and classify marine species from underwater images and videos. For instance, AI is being used to monitor coral reefs for signs of bleaching or disease and to track the populations of fish and endangered species. This technology can reduce the human effort required for image analysis by a significant margin.
Acoustic monitoring: AI analyzes soundscapes collected by hydrophones to identify, localize, and track vocalizing marine animals like whales and dolphins. Machine learning algorithms can distinguish the calls of different species or even individual animals, providing crucial insights into their behavior, migration patterns, and communication. Projects like SurfPerch, developed by Google Research and DeepMind, have used AI to create bioacoustic libraries that assess coral reef health based on ambient noise levels.