Subject:
|
Re: Human Mood Dection (Through Voice)
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 8 Dec 1999 17:25:04 GMT
|
Original-From:
|
Robert Ashford <rashfor@apcNOSPAM.net>
|
Viewed:
|
551 times
|
| |
|
|
That sounds interesting, but I don't think it would be practical to make. I
think there is a lot of processing power required to interpret the stress in
ones voice. The software/hardware lie detectors that I have seen nust have
a way to know what is a truthfull tone of voice before being able to detect
an untruthfull tone. ie you ask the person his/her name, and a bunch of
other questions that will give a truthfull or response and let the software
know what toneal range is truthfull.
It may be easyer to use ether the change in temporature or the change in
resistance of the human skin to detect mood. These are some of the inputs a
polygraph uses along with heart rate and respiration to determine stress.
The processing could be done on the RCX with this type of system.
Robert Ashford
www.apc.net/rashfor/robotics/
> Subject: Human Mood Dection (Through Voice)
> Greetings All,
>
> (I'm new to the Mindstorms, so bare with me if this has been done • already).
>
> Does anyone think this is feasible? (This could be implemented as a
> sensor). Basically a voice "reader" that could be used to determine
> a persons mood, or even for voice recognition.
>
> The first step is not to difficult. A 20$ Digital Voice recorder can • easily
> be acquired from Radio Shack.
>
> The second step, would be the tough part. As far as on a computer end,
> one could do this. There's many lie-detection software programs out there
> (whether they work or not is a different issue, but they do tell quite • well
> if there is Tension in the voice - which is all that we would need). Plus
> there
> is voice detection software I've used that I never needs to be trained.
> But,
> once again, we don't need complete recognition.
>
> I think this could be implemented in a separate unit, which would then
> output RAW data (to not kill the Lego CPU's processing power). This
> shouldn't be to hard - a mood detection algorithm wouldn't be terribly
> difficult. Just off the top of my head, I guess the algorithm could • search
> for
> - sudden changes in pitch
> - jittery speech
> - pauses
> But, I'm sure Lie detection and Mood is well documented and could be • easily
> researched.
>
> If anyone knows of a project like this already in the making, please let • me
> know.
> Otherwise, I would definitely be interested giving a whirl at something • like
> this.
> The only question is the feasibility of it ....
>
>
>
> Alex p
|
|
Message is in Reply To:
 | | Human Mood Dection (Through Voice)
|
| Greetings All, (I'm new to the Mindstorms, so bare with me if this has been done already). Does anyone think this is feasible? (This could be implemented as a sensor). Basically a voice "reader" that could be used to determine a persons mood, or (...) (25 years ago, 5-Dec-99, to lugnet.robotics)
|
4 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
