Can AI models trained on human speech help us understand dogs?
(2024/06/06)
- Reference: 1717663028
- News link: https://www.theregister.co.uk/2024/06/06/dog_speech_ai_models/
- Source link:
People who want to understand their dogs might be about to be given a helping paw by AI, according to the latest study.
Researchers from the University of Michigan are developing tools that can identify whether a dog’s bark conveys playfulness or aggression, which should already be apparent to many a Reg reader.
ML suggests all that relaxing whale song might just be human-esque gossiping [1]READ MORE
At the same time they hope to understand other information from animal vocalizations, such as the animal’s age, breed and sex. Working with Mexico’s National Institute of Astrophysics, Optics and Electronics Institute, the Michigan team found AI models originally trained on human speech can be used as a starting point to train new systems that target animal communication.
The results were presented at the Joint International Conference on Computational Linguistics, Language Resources and Evaluation.
“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, Collegiate Professor of Computer Science and Engineering, and Janice M. Jenkins, director of U-M’s AI Laboratory which carried out the work.
[2]
“There is so much we don’t yet know about the animals that share this world with us. 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,” she said.
[3]
[4]
But while human language AI models are trained on a huge corpus of written text, dogs are less well known for typing, and their voices are recorded less often than humans.
To overcome the problem, the researchers are repurposing an existing model that was originally designed to analyze human speech. The foundation from various voice-enabled technologies has been trained to pick out important features of human speech, such as tone, pitch, and accent.
[5]Airbus shows off uncrewedAI-powered Wingman for fighter pilots
[6]AI Catholic 'priest' defrocked after recommending Gatorade baptism
[7]Some scientists can't stop using AI to write research papers
[8]Boston Dynamics' humanoid Atlas is dead, long live the ... new commercial Atlas
“These models are able to learn and encode the incredibly complex patterns of human language and speech,” said Artem Abzaliev, lead author and doctoral student.
With the human speech model Wav2Vec2, researchers built a dataset of dog vocalizations recorded from 74 dogs of varying breed, age and sex, in a variety of contexts.
[9]
They found that Wav2Vec2 improved on other models trained specifically on dog bark data — with accuracy figures up to 70 percent — and succeeded at four classification tasks.
How is dog bark data gathered, you ask? The researchers record dog barks from a number of situations: playing, aggression etc. Then they test the model's ability to identify bark examples: very aggressive barking at a stranger; normal barking at a stranger; negative squeal; negative grunt (in the presence of a stranger). So the researchers know which is which because they already know the context of the barks.
"This is the first time that techniques optimized for human speech have been built upon to help with the decoding of animal communication,” Mihalcea said. “Our results show that the sounds and patterns derived from human speech can serve as a foundation for analyzing and understanding the acoustic patterns of other sounds, such as animal vocalizations.” ®
Get our [10]Tech Resources
[1] https://www.theregister.com/2024/05/09/ai_whale_language/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2024/06/05/airbus_wingman_drone/
[6] https://www.theregister.com/2024/05/03/ai_catholic_priest/
[7] https://www.theregister.com/2024/05/03/ai_scientific_articles/
[8] https://www.theregister.com/2024/04/17/boston_dynamics_atlas_is_dead/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://whitepapers.theregister.com/
Researchers from the University of Michigan are developing tools that can identify whether a dog’s bark conveys playfulness or aggression, which should already be apparent to many a Reg reader.
ML suggests all that relaxing whale song might just be human-esque gossiping [1]READ MORE
At the same time they hope to understand other information from animal vocalizations, such as the animal’s age, breed and sex. Working with Mexico’s National Institute of Astrophysics, Optics and Electronics Institute, the Michigan team found AI models originally trained on human speech can be used as a starting point to train new systems that target animal communication.
The results were presented at the Joint International Conference on Computational Linguistics, Language Resources and Evaluation.
“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, Collegiate Professor of Computer Science and Engineering, and Janice M. Jenkins, director of U-M’s AI Laboratory which carried out the work.
[2]
“There is so much we don’t yet know about the animals that share this world with us. 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,” she said.
[3]
[4]
But while human language AI models are trained on a huge corpus of written text, dogs are less well known for typing, and their voices are recorded less often than humans.
To overcome the problem, the researchers are repurposing an existing model that was originally designed to analyze human speech. The foundation from various voice-enabled technologies has been trained to pick out important features of human speech, such as tone, pitch, and accent.
[5]Airbus shows off uncrewedAI-powered Wingman for fighter pilots
[6]AI Catholic 'priest' defrocked after recommending Gatorade baptism
[7]Some scientists can't stop using AI to write research papers
[8]Boston Dynamics' humanoid Atlas is dead, long live the ... new commercial Atlas
“These models are able to learn and encode the incredibly complex patterns of human language and speech,” said Artem Abzaliev, lead author and doctoral student.
With the human speech model Wav2Vec2, researchers built a dataset of dog vocalizations recorded from 74 dogs of varying breed, age and sex, in a variety of contexts.
[9]
They found that Wav2Vec2 improved on other models trained specifically on dog bark data — with accuracy figures up to 70 percent — and succeeded at four classification tasks.
How is dog bark data gathered, you ask? The researchers record dog barks from a number of situations: playing, aggression etc. Then they test the model's ability to identify bark examples: very aggressive barking at a stranger; normal barking at a stranger; negative squeal; negative grunt (in the presence of a stranger). So the researchers know which is which because they already know the context of the barks.
"This is the first time that techniques optimized for human speech have been built upon to help with the decoding of animal communication,” Mihalcea said. “Our results show that the sounds and patterns derived from human speech can serve as a foundation for analyzing and understanding the acoustic patterns of other sounds, such as animal vocalizations.” ®
Get our [10]Tech Resources
[1] https://www.theregister.com/2024/05/09/ai_whale_language/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2024/06/05/airbus_wingman_drone/
[6] https://www.theregister.com/2024/05/03/ai_catholic_priest/
[7] https://www.theregister.com/2024/05/03/ai_scientific_articles/
[8] https://www.theregister.com/2024/04/17/boston_dynamics_atlas_is_dead/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmGIwnqCXA7nJD2VbyFwvQAAAJQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://whitepapers.theregister.com/
Re: Grant capture 'science'?
Pascal Monett
Agreed. Even with humans, there are many silent forms of expression that no AI can pick up through language. I mean, just ask any married man who feels the need to ask his wife if everything is okay.
Animals have a multitude of physical options to express themselves as well. A cat doesn't need to hiss at you to make you understand that you're unwelcome.
But, this is a start. However ridiculous it might seem at first glance, it might provide a few clues that can be built upon in the coming years. I say go for it. Nobody knows what we might find out in the long run.
So...
Anonymous Coward
They intend to get AI one stop beyond barking?
No
m4r35n357
See above
Except ...
jake
... Dogs don't use vocalization as a primary communication tool.
However, it's a novel bit of research that makes use of one of today's largest grant money cash cows. The researchers might be able to milk it for a year or three, until people realize we've been in the current AI Winter for a couple years already.
Grant capture 'science'?
I see several glaring flaws in this from the scientific perspective, the most prominent being:
[1] barking is only one of many ways dogs express themselves, many of which are indeed silent behaviours that would all need to be assessed;
[2] we have no real idea of what goes on in dogs' minds except inferentially and in the crudest terms (" playfulness or aggression "). I doubt anyone would be interested in researching human responses at that primitive level;
[3] the sample is tiny -- " With the human speech model Wav2Vec2, researchers built a dataset of dog vocalizations recorded from 74 dogs of varying breed, age and sex, in a variety of contexts ".
I have a horrible feeling that this could turn out to be a " what can we apply AI to so we can land a research grant " proposal.