The LOBSTgER initiative, a collaborative effort between MIT Sea Grant, underwater photographer Keith Ellenbogen, and MIT mechanical engineering PhD student Andreas Mentzelopoulos, merges ar…
The LOBSTgER initiative, a collaborative effort between MIT Sea Grant, underwater photographer Keith Ellenbogen, and MIT mechanical engineering PhD student Andreas Mentzelopoulos, merges artificial intelligence with underwater photography to enhance scientific storytelling about marine life.
The project leverages generative AI models trained on Ellenbogen's extensive collection of underwater photographs, creating a visual archive that is both artistically compelling and ecologically relevant. This approach aims to document the biodiversity of the Gulf of Maine, which is facing rapid environmental changes, and communicate these changes to the public in novel ways.
The LOBSTgER project uses generative AI to expand the visual vocabulary of underwater photography. The models are trained on a curated library of Ellenbogen's original photographs, ensuring visual integrity and ecological accuracy. Custom code developed by Mentzelopoulos protects the process from biases.
The project's integration of art, science, and technology allows the team to develop innovative ways to visualize ocean life and reimagine how environmental narratives are conveyed. The project's core strength lies in its interdisciplinary approach, drawing on the visual language of photography, the observational rigor of marine science, and the computational power of generative AI.
This approach allows the team to create detailed and accurate images of marine life, which can then be used to educate and engage the public. The team hopes to deepen the public's connection to the natural world. The project's foundation is built upon Ellenbogen's work documenting New England's marine life.
The resulting dataset of underwater images provides the necessary data for training LOBSTgER's generative AI models. The images capture diverse angles, lighting conditions, and animal behaviors, resulting in a visual archive that is both artistically striking and biologically accurate.
The project aims to reveal hidden ocean worlds and inspire a deeper appreciation for marine ecosystems.