SouthernWorldwide.com – Google is embarking on an ambitious project to decipher the complex communication of dolphins, leveraging the power of artificial intelligence.
Dolphins have long been recognized for their remarkable intelligence and sophisticated social interactions, captivating humans for millennia.
Now, the tech giant is employing AI to understand the intricate ways dolphins communicate with each other. The ultimate goal is to enable humans to converse with these intelligent marine creatures through advanced technology.
Google has partnered with researchers from the Georgia Institute of Technology and the Wild Dolphin Project (WDP), a Florida-based non-profit organization with a 40-year history of studying and recording dolphin vocalizations. Together, they are developing a new AI model named DolphinGemma.
For decades, the WDP has meticulously linked specific sound types to particular behaviors. For instance, signature whistles are used by mother and calf dolphins for identification and reunion, while sharp, pulsed “squawks” are often heard during aggressive encounters, according to a Google blog post detailing the project.
Distinctive “click” sounds are frequently observed during courtship rituals or when dolphins are hunting sharks.
Utilizing the extensive data collected by the WDP, Google has created DolphinGemma. This model is built upon Google’s own lightweight open-source AI model, known as Gemma.
DolphinGemma has been trained to analyze a vast repository of dolphin sound recordings, with the aim of identifying patterns, structures, and potentially even the underlying meanings in their communications.
Read more : Soviet-era spacecraft returns to Earth after 53 years
The AI’s objective is to categorize these dolphin sounds, much like how human language is organized into words, sentences, and expressions.
“By identifying recurring sound patterns, clusters, and reliable sequences, the model can help researchers uncover hidden structures and potential meanings within the dolphins’ natural communication — a task previously requiring immense human effort,” states the Google blog post about the project.
“Eventually, these patterns, augmented with synthetic sounds created by the researchers to refer to objects with which the dolphins like to play, may establish a shared vocabulary with the dolphins for interactive communication.”
DolphinGemma utilizes Google’s Pixel phone technology, specifically the advanced audio recording capabilities found in Pixel devices. This technology is crucial for capturing clear, high-quality recordings of dolphin vocalizations.
The Pixel phone’s technology can effectively isolate dolphin clicks and whistles from ambient underwater noise, such as waves, boat engines, or general static. This pristine audio quality is paramount for AI models like DolphinGemma, as noisy or corrupted data can significantly hinder the AI’s learning process, researchers explained.
Google intends to release DolphinGemma as an open-source model this summer. This will allow researchers worldwide to access, utilize, and further develop the technology.
While DolphinGemma is currently trained on data from Atlantic spotted dolphins, researchers believe that with some adjustments, it could also be applied to study other dolphin species, such as bottlenose or spinner dolphins.
“By providing tools like DolphinGemma, we hope to give researchers worldwide the tools to mine their own acoustic datasets, accelerate the search for patterns and collectively deepen our understanding of these intelligent marine mammals,” the blog post concludes.
