ANN ARBOR, Mich. – Researchers at the University of Michigan are studying how artificial intelligence technology can be used to understand a dog’s communication.
Scientists are developing tools to distinguish whether a dog’s bark shows playfulness or aggression. They also want to gather more information from the dog’s vocalizations, such as its age, breed and sex.
“By using speech processing models initially trained on human speech, our research opens a new window into how we can leverage what we built so far in speech processing to start understanding the nuances of dog barks,” said Rada Mihalcea, director of U-M’s AI laboratory.
Researchers studied vocalizations from 74 dogs in different contexts.
The study found that AI models originally trained on human speech can be used as a starting point to train new systems that target animal communication.
“There is so much we don’t yet know about the animals that share this world with us,” Mihalcea adds. “Advances in AI can be used to revolutionize our understanding of animal communication, and our findings suggest that we may not have to start from scratch.”
Mihalcea said their findings indicate that the sounds and patterns originating from human speech can form the basis for examining and comprehending the acoustic patterns of other sounds, including animal vocalizations.
However, developing AI to analyze animal vocalizations is hindered by the lack of publicly available data, researchers said.
“Animal vocalizations are logistically much harder to solicit and record,” said Artem Abzaliev, lead author of the study. “They must be passively recorded in the wild or, in the case of domestic pets, with the permission of owners.”
Scientists were able to repurpose a human speech analysis model that enabled them to tap into robust voice-enabled technologies that recognize spoken words and speakers.
“These models are able to learn and encode the incredibly complex patterns of human language and speech,” Abzaliev notes. “We wanted to see if we could leverage this ability to discern and interpret dog barks.”
This new research also has important implications for animal welfare, according to the study. Understanding dog vocalizations could greatly improve how humans interpret and respond to the needs of dogs, enhancing their care and preventing dangerous situations.
