The model relies on integrating advanced technologies in the field of artificial intelligence, such as deep learning, large language models, computer vision, as well as cognitive reasoning.
“DBTH-GPT” processes a variety of data, including video recordings, terrain information, hydrodynamics, sediments, and bioacoustic signals in the oceans. Xinhua News Agency hopes that this model will contribute to moving deep-sea research from traditional analysis methods to a new stage based on intelligent interpretation and prediction.
This model has already been used to build an intelligent cognitive system specializing in the study of seamounts and deep-sea hydrothermal vent areas, under the supervision of the Second Institute of Oceanography of the Chinese Ministry of Natural Resources. This achievement falls under the umbrella of the “Digital Deep” project, which is part of the “United Nations Decade of Ocean Science.”
In future stages, this model will be available to research institutions and international organizations, with the aim of developing cognitive systems covering different deep-sea environments, such as seamounts, abyssal plains, and continental slopes.
