Computer Speaks Dog?
According to studies done at the Eötvös Loránd University in Hungary, a new software can classify barks according to different situations. It can even connect individual barks to individual dogs. I think I can distinguish Stranger’s bark from the neighborhood dogs, but with a greater pool of unknown dogs, I don’t know if I would be able to pinpoint his unique set of woofs.
The software analyzed over 6000 barks from 14 willingly vocal Hungarian sheepdogs in six different situations: “stranger,” “fight,” “walk,” “alone,” “ball,” and “play.” The barks were taped and then transferred to a computer, where they were digitalized and individual barks were coded, classified, and evaluated.
The software correctly classified the barks 43 percent of the time, with recognition rates being highest for “fight” and “stranger.”
The software’s algorithm was 52 % accurate in matching up the individual barks of the Hungarian sheepdogs.
The scientists conclude, “The use of advanced machine learning algorithms to classify and analyze animal sounds opens new perspectives for the understanding of animal communication… The promising results obtained strongly suggest that advanced machine learning approaches deserve to be considered as a new relevant tool for ethology*.”
* Ethology: the study of animal behavior, with a focus on behavioral patterns in natural environments.
June 10th, 2008 at 10:16 pm
Hmm, that’s interesting isn’t it? I’m not sure how that is useful but it is a good reminder how dogs have several different kinds of barks and that their sounds do mean something. Of course, most of their communication is through energy and body language. Sometimes I think humans put too much emphasis on animals’ verbal communication just because that is what we rely on.
June 10th, 2008 at 11:02 pm
That is pretty interesting! I can usually pick each of my dogs’ barks out but sometimes i get them confused… would be interesting to find what a computer has to say about barks. I definitely know that different barks mean different things. Levi barks at the fridge top when he wants the treats up there!
June 11th, 2008 at 2:39 pm
There is no question I can recognize Daisy’s bark. I can not discern the same number of variations as tested in the study. I know fear or stranger approaching. I also not play but that is about it.
However I am not sure the study proved much of anything. At 43% you can get better results by tossing a coin. The results could just be luck. While intuitively I agree with the premise. There is no question dogs’ bark change with a given situation. do not think the study showed the conclusion they claim. They need to work on their program.
June 11th, 2008 at 9:17 pm
That’s quite interesting. I think they are going to have to get alot more accurate before I shell out for it. I think as we bond with out dogs and spend time with them we are able to distinguish their barks. I can tell what hungry, outside and someone’s here sounds like for all 6 of mine.