The LOBSTgER initiative, a collaborative project between MIT Sea Grant, Keith Ellenbogen (underwater photographer), and Andreas Mentzelopoulos (MIT PhD student), merges artificial intellige…
The LOBSTgER initiative, a collaborative project between MIT Sea Grant, Keith Ellenbogen (underwater photographer), and Andreas Mentzelopoulos (MIT PhD student), merges artificial intelligence with underwater photography to enhance scientific storytelling about marine life in the Gulf of Maine.
The project leverages generative AI models trained on Ellenbogen's extensive collection of original underwater photographs, which meticulously document the region's diverse marine ecosystem. This approach aims to create visually compelling and ecologically relevant imagery to deepen public understanding of the changing ocean environment.
The core of LOBSTgER lies in its interdisciplinary nature, blending the artistic vision of photography, the scientific rigor of marine biology, and the computational power of generative AI. By utilizing a curated dataset of high-quality underwater images, the project ensures visual integrity and ecological accuracy in its AI-generated outputs.
Mentzelopoulos's custom code further protects the process from potential biases. This innovative approach not only offers new ways to visualize ocean life but also redefines how environmental stories are communicated. The initiative faces the inherent challenges of underwater photography in the Gulf of Maine, such as limited visibility and the unpredictable movement of marine life.
Ellenbogen's previous work, "Space to Sea: Visualizing New England's Ocean Wilderness," provides the foundational dataset for training the AI models. The resulting visual archive, capturing diverse angles, lighting conditions, and animal behaviors, serves as a valuable resource for both artistic expression and scientific research.
In essence, LOBSTgER represents a creative experiment and a research tool, reflecting MIT's commitment to interdisciplinary innovation. By synthesizing art, science, and technology, the project strives to generate a deeper public connection to the natural world. The project's use of generative AI promises to expand the visual vocabulary for marine conservation, offering new perspectives on the beauty and fragility of the ocean environment.