AI could soon translate your dog’s barks


Scientists believe that dog lovers could soon use AI to decode what those howls and whines from their furry friends really mean.

Scientists at the University of Michigan found that AI models originally trained on human speech can be used as a starting point to train new systems that target canine communication.

According to the researchers, while AI application in animal language translation showed positive results, the main obstacle is lack of publicly available data. In contrast, there is plenty of data with human speech.

“Animal vocalizations are logistically much harder to solicit and record,” said Artem Abzaliev, a research lead and doctoral student in computer science and engineering.

“They must be passively recorded in the wild or, in the case of domestic pets, with the permission of owners.”

The researchers used a dataset of dog vocalizations recorded from 74 dogs of varying breeds, age and sex, in a variety of contexts to train a speech representation model called Wav2Vec2.

The AI model managed to successfully identify the breed, sex, and age of dogs by analyzing their vocalizations. It also accurately interpreted the context of the barks, distinguishing between playful and aggressive sounds.

The researchers discovered that their model outperformed other models specifically trained on dog bark data, achieving accuracy rates of up to 70%.

The scientists believe that their discovery could contribute to a better understanding of animals and their welfare, in addition to the further application of AI in the field of biology.

Discovered animal alphabet

AI is leading the way in animal research. In early May, scientists at MIT's Computer Science and Artificial Intelligence Lab used AI models to analyze the sound of sperm whales.

The research showed that their calls bear a striking resemblance to human language, the first ever noticed in the animal kingdom.

Analysis of whale communication showed that the clicks made by sea mammals are not random but have a structure and depend on the context, which allowed scientists to categorize the sounds into a whale alphabet.

While AI helped in this staggering discovery, it remains unknown what whales are actually saying.


More from Cybernews:

Russian cyber operations are largest threat to Olympics, Google warns

Too many cookies: Chrome to reduce the load for annoyed users

IKEA opens its first store in the Roblox universe

Stop using the Humane AI Pin charger: it might catch fire

US regulators roll up sleeves for Microsoft, OpenAI, and Nvidia inquiries