Speaker verification
-
I need to create a speaker verification software in c#.Is there any API like SAPI to do this? i googled for a while and i only get licenced DLL(800$). is there any article written about speaker verification in .net
Play a WAVE file that says, "if you can hear this, press RETURN". :-D
-
Play a WAVE file that says, "if you can hear this, press RETURN". :-D
-
is zat suppose 2 be a jok or wat? if u dont know wat 2 say plz dont say nothin at all :sigh:
-
I need to create a speaker verification software in c#.Is there any API like SAPI to do this? i googled for a while and i only get licenced DLL(800$). is there any article written about speaker verification in .net
Here's one approach: 1. Collect several samples of a spoken sentence from a good-sized group of people. The sentence should have many vowel sounds. 2. Apply a fast Forrier transform to the wave samples to get amplitudes at n different frequencies. 3. Consider each set of n samples to be a vector in an n-dimensional space. Compute the average vector for each person. 4. Compute the Euclidean distance among the average vectors for all the people, and take the smallest distance, d. 5. To verify a speaker, get the vector for the sample sentence from the speaker, and see if it's within d/2 of the average vector for the person you're trying to verify. If it is, assume it's the same speaker.
-
I need to create a speaker verification software in c#.Is there any API like SAPI to do this? i googled for a while and i only get licenced DLL(800$). is there any article written about speaker verification in .net
You have to be more specific about what you mean. I have seen "speaker verification" that means they use voice recognition to verify who the user is who is speaking. But if you mean that you want to verify that there are devices attached to output sound that is a different matter (and probably impossible unless you mean the internal PC speak er used for POST beeps).
-
Here's one approach: 1. Collect several samples of a spoken sentence from a good-sized group of people. The sentence should have many vowel sounds. 2. Apply a fast Forrier transform to the wave samples to get amplitudes at n different frequencies. 3. Consider each set of n samples to be a vector in an n-dimensional space. Compute the average vector for each person. 4. Compute the Euclidean distance among the average vectors for all the people, and take the smallest distance, d. 5. To verify a speaker, get the vector for the sample sentence from the speaker, and see if it's within d/2 of the average vector for the person you're trying to verify. If it is, assume it's the same speaker.
That might be good for microphone verification.
-
is zat suppose 2 be a jok or wat? if u dont know wat 2 say plz dont say nothin at all :sigh:
Are you supposed to be a programmer, or what? If you don't have the maturity to recognize an answer as being viable, don't respond to it. There's no way to programatically detect whether or not a speaker is plugged in, nevermind trying to detect if they're turned on. If you had any coding skills, you MIGHT be able to do it on USB speakers, but definitely NOT when something is plugged into a RCA jack on a sound card.
.45 ACP - because shooting twice is just silly
-----
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"The staggering layers of obscenity in your statement make it a work of art on so many levels." - J. Jystad, 2001 -
Yes. Note the joke icon.
I thought it was a good answer, and thought his response was crap. Maybe my tolerance is kinda low after having seen some of the Quick Answer questions this morning.
.45 ACP - because shooting twice is just silly
-----
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"The staggering layers of obscenity in your statement make it a work of art on so many levels." - J. Jystad, 2001 -
Are you supposed to be a programmer, or what? If you don't have the maturity to recognize an answer as being viable, don't respond to it. There's no way to programatically detect whether or not a speaker is plugged in, nevermind trying to detect if they're turned on. If you had any coding skills, you MIGHT be able to do it on USB speakers, but definitely NOT when something is plugged into a RCA jack on a sound card.
.45 ACP - because shooting twice is just silly
-----
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"The staggering layers of obscenity in your statement make it a work of art on so many levels." - J. Jystad, 2001 -
You have to be more specific about what you mean. I have seen "speaker verification" that means they use voice recognition to verify who the user is who is speaking. But if you mean that you want to verify that there are devices attached to output sound that is a different matter (and probably impossible unless you mean the internal PC speak er used for POST beeps).
-
Are you supposed to be a programmer, or what? If you don't have the maturity to recognize an answer as being viable, don't respond to it. There's no way to programatically detect whether or not a speaker is plugged in, nevermind trying to detect if they're turned on. If you had any coding skills, you MIGHT be able to do it on USB speakers, but definitely NOT when something is plugged into a RCA jack on a sound card.
.45 ACP - because shooting twice is just silly
-----
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"The staggering layers of obscenity in your statement make it a work of art on so many levels." - J. Jystad, 2001 -
OK PROGRAMMER.....Speaker verification means voice recognition... :-D if u don't understand voice recognition means recognizing who is speaking by analyzing his voice... if u still don't get it GOOGLE on "Speaker verification" :laugh:
If you meant "voice recognition", you should have said "voice recognition". Dipshit...
.45 ACP - because shooting twice is just silly
-----
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"The staggering layers of obscenity in your statement make it a work of art on so many levels." - J. Jystad, 2001 -
I need to create a speaker verification software in c#.Is there any API like SAPI to do this? i googled for a while and i only get licenced DLL(800$). is there any article written about speaker verification in .net
The System.Speech.Recognition engine is used to recognize a user's voice and convert it into text. The SAPI 5.3 recognition engine supports the W3C standard -- Speech Recognition Grammar Specification (SRGS), a markup language that defines how and what words are recognized, and also added support for Semantic Interpretation. take a look at this[^] article for more details.
