Rick Lyons <R.Lyons@_bogus_ieee.org> wrote:> On 05 Aug 2010 11:45:47 GMT, Ico <usenet@zeev.nl> wrote: > >>A potential customer requested a feasibiltiy study for a product >>requiring detecion of dog barks in a real-time recorderd digital audio >>signal. Functional requirements describe: >> >>"Barks should be detected with a accuracy of at least 80%, while other >>(possibly loud) signals like traffic, playing children, etc should >>generate false positives in no more then 10% of the cases." >> >>Any tips on algorithms or literature to get me started ? > > Hello Ico, Wow, that sounds like an interesting project. Your > problem, I believe, falls in the category of "signal recognition" (a > topic of which I'm shamefully ignorant). I wonder if some of the > techniques used for human "voice recognition" might be applicable to > your problem. My guess is that your "barking detection" problem is > NOT an easy problem to solve.I'm afraid so, although I think there are some interesting properties about barks to start with, see my other post from a few minutes ago.> Concerning the barking of my neighbors' dogs, I'd be willing to work > on this project for free if it included: (1) detection of dog > barking, and (2) upon detection, application of a severe electric > shock to the rectums of my human neighbors.Actually, I am extra motivated by this project for personal reasons as well. I'm not sure about the electroshocks yet, but a nice start would be to be able to hand my neighbours a *very* detailed record of all barks over a few weeks, to make them see their cute little doggy does not "Bark only once in a while, and not at night at all"
Detecting dog barks
Started by ●August 5, 2010
Reply by ●August 5, 20102010-08-05
Reply by ●August 5, 20102010-08-05
Ico wrote:> A potential customer requested a feasibiltiy study for a product > requiring detecion of dog barks in a real-time recorderd digital audio > signal. Functional requirements describe: > > "Barks should be detected with a accuracy of at least 80%, while other > (possibly loud) signals like traffic, playing children, etc should > generate false positives in no more then 10% of the cases." > > Any tips on algorithms or literature to get me started ?The key to detection could be every next bark is similar to the previous bark. You have to detect periodic bursts with specific mix of voiced and unvoiced energy. There is a ton of literature on speech decoding; you need to read on feature detection. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
Reply by ●August 5, 20102010-08-05
On 8/5/2010 10:34 AM, Ico wrote:> Rick Lyons<R.Lyons@_bogus_ieee.org> wrote: >> On 05 Aug 2010 11:45:47 GMT, Ico<usenet@zeev.nl> wrote: >> >>> A potential customer requested a feasibiltiy study for a product >>> requiring detecion of dog barks in a real-time recorderd digital audio >>> signal. Functional requirements describe: >>> >>> "Barks should be detected with a accuracy of at least 80%, while other >>> (possibly loud) signals like traffic, playing children, etc should >>> generate false positives in no more then 10% of the cases." >>> >>> Any tips on algorithms or literature to get me started ? >> >> Hello Ico, Wow, that sounds like an interesting project. Your >> problem, I believe, falls in the category of "signal recognition" (a >> topic of which I'm shamefully ignorant). I wonder if some of the >> techniques used for human "voice recognition" might be applicable to >> your problem. My guess is that your "barking detection" problem is >> NOT an easy problem to solve. > > I'm afraid so, although I think there are some interesting properties > about barks to start with, see my other post from a few minutes ago. > >> Concerning the barking of my neighbors' dogs, I'd be willing to work >> on this project for free if it included: (1) detection of dog >> barking, and (2) upon detection, application of a severe electric >> shock to the rectums of my human neighbors. > > Actually, I am extra motivated by this project for personal reasons as > well. I'm not sure about the electroshocks yet, but a nice start would > be to be able to hand my neighbours a *very* detailed record of all > barks over a few weeks, to make them see their cute little doggy does > not "Bark only once in a while, and not at night at all"The neighbors won't likely believe the output of a bark detector. Use a two-track audio recorder, one track tor timestamps, the other to record the barks themselves, (The timestamps could be on the same track as the audio, with a small readout for them.) the whole thing sound operated. A suitable delay line will allow recording of the barks' beginnings. The scheme has the advantage that the barking will be nearly continuous upon playback, intensifying the effect. Jerry -- "Tell me what you need and I'll show you how to do without it." �����������������������������������������������������������������������
Reply by ●August 5, 20102010-08-05
Clay <clay@claysturner.com> wrote: (snip)> Simple, get your own dog and attach a sound detector with a high > detection threshold to your dog. When your dog hears the distant dog's > barks, your dog will act as a bark repeater/amplifier and set off your > detector which will be rarely falsed because of the high threshold of > detection.I bionic bark detector! -- glen
Reply by ●August 5, 20102010-08-05
>Rick Lyons <R.Lyons@_bogus_ieee.org> wrote: >> On 05 Aug 2010 11:45:47 GMT, Ico <usenet@zeev.nl> wrote: >> >>>A potential customer requested a feasibiltiy study for a product >>>requiring detecion of dog barks in a real-time recorderd digital audio >>>signal. Functional requirements describe: >>> >>>"Barks should be detected with a accuracy of at least 80%, while other >>>(possibly loud) signals like traffic, playing children, etc should >>>generate false positives in no more then 10% of the cases." >>> >>>Any tips on algorithms or literature to get me started ? >> >> Hello Ico, Wow, that sounds like an interesting project. Your >> problem, I believe, falls in the category of "signal recognition" (a >> topic of which I'm shamefully ignorant). I wonder if some of the >> techniques used for human "voice recognition" might be applicable to >> your problem. My guess is that your "barking detection" problem is >> NOT an easy problem to solve. > >I'm afraid so, although I think there are some interesting properties >about barks to start with, see my other post from a few minutes ago. > >> Concerning the barking of my neighbors' dogs, I'd be willing to work >> on this project for free if it included: (1) detection of dog >> barking, and (2) upon detection, application of a severe electric >> shock to the rectums of my human neighbors. > >Actually, I am extra motivated by this project for personal reasons as >well. I'm not sure about the electroshocks yet, but a nice start would >be to be able to hand my neighbours a *very* detailed record of all >barks over a few weeks, to make them see their cute little doggy does >not "Bark only once in a while, and not at night at all" > >There are numerous consumer products available that detect dog barks, then responds with ultrasound that the dog quickly associates with its own barking (allegedly). Worth looking into for this project and your "personal reasons" ;) Bryan
Reply by ●August 5, 20102010-08-05
Ico wrote:> A potential customer requested a feasibiltiy study for a product > requiring detecion of dog barks in a real-time recorderd digital audio > signal. Functional requirements describe: > > "Barks should be detected with a accuracy of at least 80%, while other > (possibly loud) signals like traffic, playing children, etc should > generate false positives in no more then 10% of the cases." > > Any tips on algorithms or literature to get me started ? > > Thanks, > > Ico >If you did a "fair" job of detection, accepting more false alerts in trade for fewer false rests. I'm assuming that you're going to be wanting to detect multiple-bark sequences more than single barks. saved audio segments that are "likely", then edit out the false alerted segments manually (and, perhaps eventually automatically), You'd have a short(er) audio record of all the dog barks. I figure that the false alerts would have *some* noticeable difference that might be automated once you have the "rough" data. And, I'm assuming that the composite / compressed audio record doesn't have to be generated in real time. The problem is likely tougher if it has to run in real time. Rick's shock machine would have to run in real time to be effective I should think. Even Pavlov would say so, I'm sure. I would have mentioned the off-the-shelf products but I'll bet they operate on amplitude more than anything - so have to have high SNR and, thus, close proximity. Fred
Reply by ●August 5, 20102010-08-05
On Aug 5, 1:26=A0pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:> Clay <c...@claysturner.com> wrote: > > (snip) > > > Simple, get your own dog and attach a sound detector with a high > > detection threshold to your dog. When your dog hears the distant dog's > > barks, your dog will act as a bark repeater/amplifier and set off your > > detector which will be rarely falsed because of the high threshold of > > detection. > > I bionic bark detector! > > -- glenThe problem I'm still trying to figure out is why dogs don't bark themselves deaf? I've asked veternarians this, and the ones I've asked don't know. I wonder if the dog's ear canals pinch close during barking? Clay
Reply by ●August 5, 20102010-08-05
On Thu, 5 Aug 2010 12:14:09 -0700 (PDT), Clay <clay@claysturner.com> wrote:>On Aug 5, 1:26=A0pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote: >> Clay <c...@claysturner.com> wrote: >> >> (snip) >> >> > Simple, get your own dog and attach a sound detector with a high >> > detection threshold to your dog. When your dog hears the distant dog's >> > barks, your dog will act as a bark repeater/amplifier and set off your >> > detector which will be rarely falsed because of the high threshold of >> > detection. >> >> I bionic bark detector! >> >> -- glen > >The problem I'm still trying to figure out is why dogs don't bark >themselves deaf? I've asked veternarians this, and the ones I've asked >don't know. I wonder if the dog's ear canals pinch close during >barking? > >ClayAdaptive noise cancelling/line enhancing. Or something like that. I don't think a bark is loud enough to do any physical damage. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.abineau.com
Reply by ●August 5, 20102010-08-05
Eric Jacobsen <eric.jacobsen@ieee.org> wrote:> On Thu, 5 Aug 2010 12:14:09 -0700 (PDT), Clay <clay@claysturner.com> > wrote: > >>The problem I'm still trying to figure out is why dogs don't bark >>themselves deaf? I've asked veternarians this, and the ones I've asked >>don't know. I wonder if the dog's ear canals pinch close during >>barking? > > Adaptive noise cancelling/line enhancing. Or something like that. I > don't think a bark is loud enough to do any physical damage.I'm sure my neighbours dog is loud enough, because *I* am about to make the physical damage happen here!
