Scientists are using natural language processing and AI to decode the communication of sperm whales and potentially unlock inter-species communication, while also studying their vocalizations to better understand and care for them amidst environmental threats.
Scientists are using sperm whales' codas to test non-human translation abilities and launched the Citation Translation Initiative to collect data from whales around Dominica.
Animals communicate with each other through various means, but whether or not they use language like humans do is difficult to determine.
Sperm whales are a good candidate to test our non-human translation abilities as they use discrete sequences of clicks called codas, which researchers believe is the basic communication unit for their language and differs between whales in different regions.
SETI launched the Citation Translation Initiative to collect data from sperm whales around Dominica to understand how computers might translate whale language.
Unsupervised translation between languages is now possible using statistical properties of words, eliminating the need for a Rosetta Stone.
Machine translation uses an encoder-decoder deep learning model with two neural networks, but requires human supervision and cannot translate unknown languages.
Unsupervised translation is now possible by calculating statistical properties of words and assigning them to a point cloud, allowing for translation between languages without the need for a Rosetta Stone.
Scientists are using AI to translate sperm whale communication, but need a lot of data, so they're studying their vocalizations off the coast of Dominica for the SETI project.
Scientists are excited about the possibility of using AI to translate sperm whale communication, but it requires a large amount of data to be successful.
Scientists are studying the vocalizations of sperm whales off the coast of Dominica to gather data for the SETI project, estimating that they produce between 400 million to 4 billion clicks per year for communication purposes.
Scientists are using recording devices and autonomous robots to gather and analyze data on the sounds and behavior of sperm whales.
Autonomous and semi-autonomous devices will collect massive amounts of background whale bioacoustic data from tethered buoy arrays positioned at different depths to study patterns in the language of sperm whales.
Hydrophones were used during different stages of diving to track whale movements over space and time.
Scientists are using recording devices attached to whales to gather detailed data on their sounds and behavior, while also utilizing autonomous sound recording robots and machine learning tools to parse and analyze the data.
Researchers are using an algorithm to analyze the codas of sperm whales to categorize them by which whale made which call, but there are challenges in determining meaningful variations in the clicks.
Understanding whale language is challenging, but recognizing communication patterns can help us care for them amidst the threats of climate change, pollution, and shipping impacts.
Understanding whale language is difficult as we cannot assume they have the same emotions, sense of self, or perception of time as humans, but recognizing patterns in their communication can provide insight into their lives and help us care for them.
Cetaceans, including whales and dolphins, are facing a difficult future due to climate change, changing habitats, food availability, pollution, and impacts from the shipping industry.