Winner of a video, I been tryin to find out about "how to make a real iron man costume" for a while now, and I think this has helped. Have you ever come across - Xenacob Heyndrew Guide - (should be on google have a look ) ? Ive heard some decent things about it and my brother in law got excellent success with it.
3 years ago I built a BB-8 version, controlled with voice commands. My best video got just 837 views. You can take a look in my playlist on my channel.
@@JohnDlugosz - unlikely. The array of speech patterns is probably matched to the microphone input and it just returns the programmed response. The speech patterns are fixed.
The speech recognition unit returns a number based on the command said. This number starts at 1. Array indeces in C/C++ start at 0. So the cmd variable will return you a number between 1 and 20, and the text array needs a number between 0 and 19. That's the reason for the "cmd - 1"
Fun fact, a friend of mine used to do certification testing for the Xbox when the Kinect was coming out, and apparently saying everything with a french accent was the best way to get it to recognize your commands.
With all your talents, I am really surprised that you dont have an active live robot mascot thats part of your presentations. Similar to Starks fire extinguisher robot :)
That Cool Components Shield could be used as a programming class project, escape the room puzzle, connecting speech recognition to mp3 player to play response ("Number one", "number two", "Spanish Inquisition", "Opening Door" ) to audibly respond to your voice,....etc
As a big fan of James, I couldn't have missed the opportunity of building a BB-8 version (using Lego parts), that you can control with voice commands. I used a Geeetech voice recognition module and an Arduino nano. You can also command other RC stuff. Instructions are available on Instructables website.
I like this. I think it would've been better by you explaining speaker-dependent and speaker-independent methods. I also like your Jetson Nano ROS video #2 because I've been using Willow Garage & The Construct's tutorials but your video made a lot of ROS "just click" for me
Would be really cool if you could demo some spacial tracking products, similar in function to what VR setups use. I'm looking into different solutions to mess about with custom tracking lol.
using these products and then showing how to use with with your own projects is a very good idea especially if it helps you with sponsored products between your normal videos that can take a while to print - build ect. Just keep it relevant to what you do.
Yessssss😍😍 that's what I was waiting to make a control system by voice without using Alexa or Google or any of those assistances to not connect it with Internet, that would make it safe system no one can hack it
finally back to classic style. Although there's nothing wrong with your ongoing projects I do prefer the info burst ones you do showcasing specific tech I may one day want to use myself.
Hi James, I like these type of videos as I wouldn't really come across some of these components otherwise :) It would be neat to see more involved examples to show the possibilities but I understand there's not much reward in that
These chips are a good alternative to the software based programs in the following ways 1) They can work offline 2) They privacy friendly as they do not use popular data collection platforms 3) More compact and power friendly than running a computer, even if we compare them against SBC's like raspberry pi's
Great find with Speakup! As far as "New Products" I, and probably most, will admit to being cool gadget nut! Imagine using this in a production environment where you say your commands instead of typing on a keyboard. A keyboard that could potentially spread the flu or even something nastier.
As soon as you realize that's actually "speech recognition with output TO an Arduino (done by something else immensely more powerful and/or single-purpose-designed)" it becomes far more easy to wrap your head around.
@@AttilaAsztalos oh I get how it works, it's just crazy that it exists. Until this, you needed a massive battery or be plugged into the wall to operate your full fledged computer, even if it was a pi, to do this. Now you can just use an Arduino, and probably a couple of AA batteries.. And that still feels a bit like magic.
I didn't realize you were using an "American accent" until you said that it works better with it! You do a pretty convincing American accent though! It's funny that the American accent you use is very cowboy! lol. Also as an American, I couldn't even begin to attempt a British accent without it being laughably terrible!
Hello James, I want to make a device that can recognize when the water of my toilet is leaking water, do you think that the boards could record a water running sound and put it near the toilet so it can recognize the sound and execute a command like send a notification to your cellphone?
@@jamesbruton Possibly. The last time I looked into the subject, there was a problem with the accent causing problems with cliping in the entered voice files, but with better microphones, microphone tech, and audio processing this might have been fixed.
Interesting options. I built a project around a dedicated chip, I think from Radio Shack, back in college 25yrs ago. You could 'train' it to recognize up to 16 words which it output as 4bit BCD if memory serves. I piped that to a decoder to control 16 LEDs for the class project.
I use the devices by particle.io. You can write and flash code (over wifi) via their web IDE. They provide and API you can use to call functions or you can set them to be published and they'll hook up with IFTTT. It's been pretty great for home automation.
The full speech recognition engine requires Internet access? Products like _Dragon Simply Speaking_ have been around since the 90's, and an R-pi is far more powerful than a PC was back then.
Just a side note for anyone having issues with MRL (myrobotlab) install or boot. Just make sure you are running the correct version of JAVA for your pc and or double check you don't have multiple versions of JAVA installed for potential cross talk issues. If all goes well on that end you can just upload MRL to your arduino and away you go. That being said, of course it doesn't need to be on your arduino to run things like speechrec, just if you are connecting other pin related services such as servos etc.
If the WebKit speech only works on full fat chrome, would it be particularly difficult to stream the speech to a remote server and stream the speech back to use it on a pi
great video but supper shitty video description, tones of cut and past generic info mostly unrelated to the video, but no recaplist/link to the product mentioned = supper annoying to really make use of the info
Thanks! Super useful. I've been compiling a list of options for speech or sound input and these are all ones I'd not heard of. Incidentally there's a google API available where you can send it a text sentence and it will send you back a handy version of it parsed out into a JSON object that details things like the subject and the action etc. Saves a bit of work (or possibly creates 5 times more) but alas, not free. That said, not expensive either and it has a fair whack of free calls. Probably enough for development or experimentation. I think it was this one. cloud.google.com/natural-language although the demo doesn't seem as clear as I remember it being in the past. There's also cloud.google.com/speech-to-text which seems to work pretty well, again I think it is a subscription but plenty of free calls to play with.
This one may be of interest too (image object analysis though - not speech) you give it an image and get back JSON saying what is in it and where. cloud.google.com/vision
I see a future video where he tells opendog to sit, shake, roll over, or play dead, and it does. :D
i have a teddybear that follows orders.
Ignore me!
he need to start making catgirl waifus
Sit, shake, lay down, NOW GO FORTH AND CONQUER THE WORLD MY CHILD!
Winner of a video, I been tryin to find out about "how to make a real iron man costume" for a while now, and I think this has helped. Have you ever come across - Xenacob Heyndrew Guide - (should be on google have a look ) ? Ive heard some decent things about it and my brother in law got excellent success with it.
I think cool components and your channel is the perfect marriage :)
I think this will get good views, lots of people into Speech Recognition these days
Thanks, hope so
3 years ago I built a BB-8 version, controlled with voice commands. My best video got just 837 views.
You can take a look in my playlist on my channel.
I'm already head of my time
James, you've probably already realized this by now, but index 1 on the grove device is probably "Hi cel" xD
And, if your code ever did return a value
@@JohnDlugosz - unlikely. The array of speech patterns is probably matched to the microphone input and it just returns the programmed response. The speech patterns are fixed.
Amazing work once more, James! I wonder if you could do more collaborations with Colin again?
Maybe, it generally happens when he needs something ;-)
@@jamesbruton please collaborate!! Both of you are amazing !!
That Bed did make me laugh as I had no idea what I was looking at at first :D
The speech recognition unit returns a number based on the command said. This number starts at 1. Array indeces in C/C++ start at 0. So the cmd variable will return you a number between 1 and 20, and the text array needs a number between 0 and 19. That's the reason for the "cmd - 1"
Fun fact, a friend of mine used to do certification testing for the Xbox when the Kinect was coming out, and apparently saying everything with a french accent was the best way to get it to recognize your commands.
dude you should make robot kits every one who thinks he should make kits, thumbs up
Yey! glad to see some love for MyRobotLab.
With all your talents, I am really surprised that you dont have an active live robot mascot thats part of your presentations. Similar to Starks fire extinguisher robot :)
"¡Robot! ¡Divide five by zero!" and that's the last thing James said before the explosion.
That Cool Components Shield could be used as a programming class project, escape the room puzzle, connecting speech recognition to mp3 player to play response ("Number one", "number two", "Spanish Inquisition", "Opening Door" ) to audibly respond to your voice,....etc
As a big fan of James, I couldn't have missed the opportunity of building a BB-8 version (using Lego parts), that you can control with voice commands.
I used a Geeetech voice recognition module and an Arduino nano. You can also command other RC stuff.
Instructions are available on Instructables website.
I like this. I think it would've been better by you explaining speaker-dependent and speaker-independent methods. I also like your Jetson Nano ROS video #2 because I've been using Willow Garage & The Construct's tutorials but your video made a lot of ROS "just click" for me
Would be really cool if you could demo some spacial tracking products, similar in function to what VR setups use.
I'm looking into different solutions to mess about with custom tracking lol.
That bed shaker looks useful for other things.
using these products and then showing how to use with with your own projects is a very good idea especially if it helps you with sponsored products between your normal videos that can take a while to print - build ect. Just keep it relevant to what you do.
Can you make an advanced walking humanoid robot with speech recognition, ROS, navigation, open cv, AI,deep learning and many other interesting stuff?
LOL @ 6:51 ... he went full Kentucky with that accent.
Been following for long time but for some reason missed the collab with Mr Furze :/ both great UA-camrs :-)
Yessssss😍😍 that's what I was waiting to make a control system by voice without using Alexa or Google or any of those assistances to not connect it with Internet, that would make it safe system no one can hack it
finally back to classic style. Although there's nothing wrong with your ongoing projects I do prefer the info burst ones you do showcasing specific tech I may one day want to use myself.
NodeRed on Raspberry pi is great for home automation projects, can do pretty much anything.
Hi James, I like these type of videos as I wouldn't really come across some of these components otherwise :)
It would be neat to see more involved examples to show the possibilities but I understand there's not much reward in that
Most of my videos are more involved projects ;-)
@@jamesbruton This is very true, you got me there lol
The most ambitious crossover ever. Great video.
Another extremely helpful video, James. Thank you!
I love your videos ❤️❤️❤️
When you were playing back audio tracks 1 to 3, I was anticipating, "Spanish Inquisition" for #3🙂
we need another two leg walking robot!!!!!!!!
With the Raspberry Pi what is the best Speech Recognition solution? I've been using one that blocks the program every now and then and is very slow.
DeepSpeech
a nerdy john oliver, subbed
I am a student from India and I got inspired to make projects from you
doesnt everyone?
@@zyugyzarc absolutely sir
This was extremely helpful pls do more stuff like this
These chips are a good alternative to the software based programs in the following ways
1) They can work offline
2) They privacy friendly as they do not use popular data collection platforms
3) More compact and power friendly than running a computer, even if we compare them against SBC's like raspberry pi's
Great video James. Good ol' myrobotlabs to save the day.
Great find with Speakup! As far as "New Products" I, and probably most, will admit to being cool gadget nut!
Imagine using this in a production environment where you say your commands instead of typing on a keyboard. A keyboard that could potentially spread the flu or even something nastier.
Speech recognition with an Arduino? Witchcraft I say. That's awesome
As soon as you realize that's actually "speech recognition with output TO an Arduino (done by something else immensely more powerful and/or single-purpose-designed)" it becomes far more easy to wrap your head around.
@@AttilaAsztalos oh I get how it works, it's just crazy that it exists. Until this, you needed a massive battery or be plugged into the wall to operate your full fledged computer, even if it was a pi, to do this. Now you can just use an Arduino, and probably a couple of AA batteries.. And that still feels a bit like magic.
Great video, very useful as well. Please upload more.
I didn't realize you were using an "American accent" until you said that it works better with it! You do a pretty convincing American accent though! It's funny that the American accent you use is very cowboy! lol. Also as an American, I couldn't even begin to attempt a British accent without it being laughably terrible!
is the click board the one you have 932-MIKROE-1534? because there is also 932-MIKROE-2375????
Hello James, I want to make a device that can recognize when the water of my toilet is leaking water, do you think that the boards could record a water running sound and put it near the toilet so it can recognize the sound and execute a command like send a notification to your cellphone?
Looks cool
I would love to see you build a Inmoov robot
Is there a way through which we can make an independent Arduino speech recognition robot with your RobotLab software ?
If you want even more simple voice recognition, check out my voice controlled power wheels
Awesome James
It would be cool to build a DIY turtle bot robot. So basically source your own components like a RPi, pi cam, lidar...
Hi , im trying to use the speak up click and having some issues with the arduino recognizing commands can you help me with that ?
But will either of these work using English with a Scottish accent?
the one where you record your own voice probably.
we need mike boyd to visit!
@@jamesbruton Possibly. The last time I looked into the subject, there was a problem with the accent causing problems with cliping in the entered voice files, but with better microphones, microphone tech, and audio processing this might have been fixed.
That is so cool, I like that video.
Interesting options.
I built a project around a dedicated chip, I think from Radio Shack, back in college 25yrs ago. You could 'train' it to recognize up to 16 words which it output as 4bit BCD if memory serves. I piped that to a decoder to control 16 LEDs for the class project.
I use the devices by particle.io. You can write and flash code (over wifi) via their web IDE. They provide and API you can use to call functions or you can set them to be published and they'll hook up with IFTTT. It's been pretty great for home automation.
I'd like to see a Raspberry pi example with Vosk. It seems to be the successor to the sphinx project.
This is amazing !!!
The full speech recognition engine requires Internet access?
Products like _Dragon Simply Speaking_ have been around since the 90's, and an R-pi is far more powerful than a PC was back then.
But Ggogle Webkit API is much better at recognising multiple people in multiple languages.
How do i run the My Robot Lav .jar file?
By having a Java Runtime Engine installed and clicking on it.
@@jamesbruton Thanks James
you generally need java 1.8, then run java -jar myrobotlab.jar
Just a side note for anyone having issues with MRL (myrobotlab) install or boot. Just make sure you are running the correct version of JAVA for your pc and or double check you don't have multiple versions of JAVA installed for potential cross talk issues.
If all goes well on that end you can just upload MRL to your arduino and away you go. That being said, of course it doesn't need to be on your arduino to run things like speechrec, just if you are connecting other pin related services such as servos etc.
Very good video
I cant find any legit video of live voicechanger, like the one you can install in a mask (smth like Dart W)) there isn't one on youtube(((
On android with code block you can use google speech to text and it work great, but require wifi or 3g / 4g.
It's - 1 because array indexes in C and C++ start with 0, not 1
I just finished the setup on a voice recognition board
Perfect
Could you make a security device that would have speech and facial recognition?
Fascinating! ✌😎🏴☠️
Love it! 👍
good video!
Now that's how to wake up like a -boss- employee...!
If the WebKit speech only works on full fat chrome, would it be particularly difficult to stream the speech to a remote server and stream the speech back to use it on a pi
no it would be fine, other than you need a remote computer
really fun video
You can use Russian EasyVR shield from Amperka
can't find the esp code
Hello sir voice control left project please help
Can you try "Pocketsphinx" speech recognition on ROS? 😆
I like it very nice
Great
"It helps if you speak with an American accent"
I suppose my British accent would sound just as funny to you 😜
Looking is good
Is my comment visible???..
I don't think so..
Hi love the vids I have to get up at 6:00 am
Proof, if proof were needed, that a Furze Safety Tie keeps you safer - much less bed-induced instability...
Nice 👍
Wow I might get this for my robot that I am making because then I won't have to press buttons on it so it can do something
I didn't even realize speech recognition has become so cheap... guess who is building a completely unnecessary room automation
This is what inspires young minds to pursue a career in science
6:49 'slight American accent' is right lol
You forgot the magic words: "Alexa..."
👍
Yo what thebheck i need to make that bed
3 minutes 👍
hi
I should be asleep
great video but supper shitty video description, tones of cut and past generic info mostly unrelated to the video, but no recaplist/link to the product mentioned = supper annoying to really make use of the info
Thanks! Super useful. I've been compiling a list of options for speech or sound input and these are all ones I'd not heard of. Incidentally there's a google API available where you can send it a text sentence and it will send you back a handy version of it parsed out into a JSON object that details things like the subject and the action etc. Saves a bit of work (or possibly creates 5 times more) but alas, not free. That said, not expensive either and it has a fair whack of free calls. Probably enough for development or experimentation. I think it was this one. cloud.google.com/natural-language although the demo doesn't seem as clear as I remember it being in the past. There's also cloud.google.com/speech-to-text which seems to work pretty well, again I think it is a subscription but plenty of free calls to play with.
This one may be of interest too (image object analysis though - not speech) you give it an image and get back JSON saying what is in it and where. cloud.google.com/vision
1:43 What is the point of... all this?
Elevyn! ua-cam.com/video/MNuFcIRlwdc/v-deo.html
First % mins
Give a like on this comment.
😁
WTF?
That is the worst American accent lol