DolphinGemma, a model trained on Atlantic spotted dolphin sounds, is set to be shared with the research community this summer. Developed through a collaboration involving the Wild Dolphin Project, Georgia Tech, and Google, this open model aims to revolutionize our understanding of cetacean communication. By making DolphinGemma available, researchers worldwide will have the opportunity to explore and analyze acoustic data from various dolphin species, potentially requiring fine-tuning for specific vocalizations. This initiative underscores the importance of collaborative efforts in scientific discovery, offering new tools to accelerate the study of marine mammals’ complex communication systems. With DolphinGemma, we are moving closer to bridging the communication gap between humans and dolphins, marking a significant step forward in marine biology research.
Read more at Google…
