ฝัง
- เผยแพร่เมื่อ 23 ก.ค. 2019
- HUSKYLENS is an easy-to-use AI vision sensor which can learn a new object, face, and color just by clicking. And it is now on Kickstarter with only US$ 19.9 for Early Bird Price!
www.kickstarter.com/projects/...
What is HUSKYLENS?
HUSKYLENS is an easy-to-use AI camera. It can learn to detect objects, faces, and colors just by clicking. The more it learns, the smarter it is.
The adoption of the new generation AI chip allows HUSKYLENS to detect faces at 30 frames per second.
HUSKYLENS can connect to Arduino, Raspberry Pi, LattePanda, or micro:bit, and make your very creative projects without playing with complex algorithms.
Please back us on Kickstarter!
www.kickstarter.com/projects/... - วิทยาศาสตร์และเทคโนโลยี
Awesome! Amazing! Terrific!
Well done, keep it up, thanks for sharing with us :)
Besides Arduino c program code, will there be some PBASIC for BS2 or PICAXE BASIC examples?
cons: it is not entirely open-source
Which is better pixy2 or this camera???
Could this lens be used to track billiard balls?
this device was 5 years into the future, if you can program it , then you may have an advantage into the future, which is now, for now>
It can take photos to an SD card but can I access those photos from the Raspberry Pi?
Hi, what is the robo project name at 0:36? Can’t find it :(
can we change the mode over arduino code. Example: i choose the face recognition then i want to change over an arduino code into the object recognition, is it possible ?
Can we include our own code into huskylens and use it,Is there any possibility
How many tags or QR codes couldn’t recognize and how many colors?
Excelente producto felicitaciones
Can the camera be extended away from the board with a longer ribbon cable?
Maybe you could send the signals through a LoRa setup
ok! when it loses the object and doesn't recognize it anymore, how to point that it's the same object so it will continue to recognize it? I wanted to buy several of these but after buying one I stuck with this problem. huge thanks for such a product btw, it's so much easier than TF. but when I can't control it, I can't use it.
You should click the learn or the button which is present on right not the button for navigation.
Then the husky lens will learn the object and starts remembering it
@@ayaansworkshop huh? i'm not talking about obvious stuff. the problem was that it learned the object but the longer i try to track the object the more probability it will stop recognising it.
Perhaps the huskylens project was ended. The huskylens forum has some few basic questions without answer, and a bunch of bugs report.
Does the Huskylens have an I2C and/or SPI interface?
Hi Walter! "Through the UART / I2C port, HuskyLens can connect to Arduino, Raspberry Pi, or micro:bit to". You can get the pdf wiki page from dfrobots web site.
@@Robots-and-androids I do not understand why everybody keeps saying use the wiki, use an Arduino or Raspberry PI when I say I'm using a PICAXE chip?
@@waltperko8389 It is not set up to use a picaxe chip. You would have to "roll your own" to get it working. It uses I2C, but I don't know if they published enough information to hack it.
@@Robots-and-androids No kidding. That is why I'm posting the question to get support. The engineers that made the HuskyLens know how to communicate with it 100% and they should be able to publish something better than just use the libraries for Arduino and Raspberry Pi etc... I don't like using "canned" software.
Would it function with a jetson nano?
Not out of the box.
A lot of show, but so far only Arduino support. Where is the support to help those of us that use PIC or PICAXE chips? So far my DFRobot blog and forum posts just get deleted ??? That is not support!
Ah, you've had the same support that I have too encountered. ;) Sorry Walter, currently no PIC or PICAXE options.
I want to learn how to create and develop projects. Do I need to learn Python or some programming language? I dont know where to start, but I have never been so fascinated in somthing, in all my life. AI, ML, Deep learning and IOT are things I want to learn and get really deep into, does anyone have any advice please? How can I do this, I do not know any programming language and am not sure what direction to go, especially with all the different dev boards such as the different ones shown which are compatible with tis product....So much different information out there Im lost on what I need to do.
Hello, first of all, I think you need to know what project to do. Secondly, do this project, study and research, or just get the results? Then, choose a suitable development board to your project. There are many common development boards, such as arduino, raspberry pie, micro:bit, lattepanda. You can find tutorials on DFRobot wiki: wiki.dfrobot.com. I think programming language is needed, such as C++, python. Hope these information will help you.
Start simple, and find a project.
Maybe you have a lawn with a sprinkler controller.... build your own.
Get a cheap relay board of eBay, an arduino to control it. Get a WIFi shield.
Get a raspberry pi, and install openhab.
Set up a mqtt client on the arduino, and control it from openhab.
Automate it using rules.
Write your own python script to control it over mqtt.
Keep on going...
Just find a project, and keep building on it.
Can you keep going, add some sensors, look at the weather, how can you predict when the sprinklers should be on, and how long. Use AI for this.
Just keep building on something you love until its amazing.
@@youyou3756 Your link leads to a page with no tutorials and the icons are not clickable. Could you check and point us to HuskyLens tutorials please.
Yes, you certainly need to learn a programming language to use this device. You will need to buy a microcontroller to use this board. I recommend the Arduino or the Microbit. I use the Arduino and if you choose that one you will need to know some elementary C language programming. On the other hand, the Microbit uses Python. There is a lot of information and online classes for both of those microcontrollers. I would stay away from the Raspberry Pi unless you are very comfortable with LINUX.
What is the difference between Huskylens and neuro cam from korea?
Hi, there! Please leave your email here, and we can have a talk about it!
@@DFROBOTCN jani-@hanmail.net
th-cam.com/video/00VxegKlK24/w-d-xo.html
There is a HUSKYLENS discord server available:
discord.gg/JwwTdu
Join our community!
How do I connect a watertight USB-endoscope-camera ??
this unit contains the camera and the ai hardware. if you already have a camera all you need is CV software on a connected computer.
I dont know why but i feel like this can make an AI VTOL with a drone
I have a husky lens and it's brilliant. I connect it to power and it learns by clicking its buttons. But then I find I can do nothing with it. The computer (an iMac) doesn't recognise it as a connected device and I have no software to control it. How can I do something with it? Can anyone guide me please?
You need a microcontroller. It communicates with a computer through a microcontroller such as Arduino, Raspberry Pi, or micro:bit. I am using an Arduino. That's pretty straight forward.
It's not a simple plug and play device for any computer l.
It is used to build prototypes like robots, self driving cars.it cannot be used as a simple device with computer you need a microcontroller such as Arduino and raspberry Pi and code it or program it.
@@ayaansworkshop Many thanks for your reply. I think I may have been over-optimistic when I bought it. It's probably a little above my abilities. I do have a Raspberry Pi 4 though so perhaps I'll be able to do something with it. I'm always optimistic! I'm going to look at your channel now.
@@Robots-and-androids Thanks for your reply. I can't believe it's two years ago that you wrote it! I did see it at the time but didn't reply. Just wanted to say thanks.
@@TikTokBrian Paul McWhorter has a best microcontroller tutorials on Internet, check that man up he has for Arduino, Raspberry, Jetson Nano, Jetson Xavier....
Have you seen? th-cam.com/video/dwUJVYEhxGM/w-d-xo.html This should be a demo with the HuskyLens to speed past a Deepracer ... maybe even get alongside one and shove it off the track!
"Making an autonomous robot has never been so easy" LOL. Making any robot that didn't come from a kit is NEVER EASY.
Iwant to sell in korea.
ديم
This is NOT a convincing demo. I hear lots about what it can do - but very little evidence of it. For example, at 1:13, is it chasing the ball? I see the ball rolling, and the robot going in more-or-less the same direction - but I was unconvinced that these were not unconnected events. The "dog" does not appear to turn at any point to follow the ball. It can apparently follow somebody riding around it on a bicycle, but that's not a very complex task. You could do it with a couple of photo transitors or LDRs.
There are better videos describing its function more in-depth. From what I gathered, the facial recognition part seems powerful in that it can recognize several faces and return coordinates of them, however there are some limitations to object identification and labeling. From what I understood, you don't get any coordinates for the objects you've trained manually, which kind of sucks (and maybe that would explain why it doesn't chase the ball?). However there are still many projects in which this is useful, so it depends on what you want really. I bought it because it's easy to use and not cluttered with a bunch of additional sensors, but still aware of the fact that I might buy a better one if I need to in the future.
That bokeh tho...
Only thing I'm worried about, is CCP involved?