Image Recognition with the Sipeed M1 Dock board

Поделиться
HTML-код
  • Опубликовано: 3 июл 2024
  • In this video am going to show a new development board, the Sipeed M1 Dock which features the revolutionary $10 K210 AI chip. Just like the ESP32, this chip is going to change everything, bringing hardware AI to the maker community.
    🛒 Sipeed M1: educ8s.tv/part/SipeedM1
    📥 Firmware: dl.sipeed.com/shareURL/MAIX/Ma...
    📥 K Flash GUI: github.com/sipeed/kflash_gui/...
    📥MaixPY IDE: dl.sipeed.com/shareURL/MAIX/Ma...
    💻 Code & Parts: educ8s.tv/sipeed-m1-dock-review
    👨‍💻 Check my new RUclips channel: bit.ly/3tku2n0
    🎮 My Android Game: bit.ly/QuizOfKnowledge
  • НаукаНаука

Комментарии • 102

  • @rodnolasco
    @rodnolasco 4 года назад +2

    This is great! Can't wait to see the future episodes. Thanks for sharing. :)

  • @rlb5261
    @rlb5261 3 года назад +1

    This is just excellent. I was able to run my Sipeed Maixduino only after watching this video. Thank you!!!!

  • @MATRIXPROEMO
    @MATRIXPROEMO 4 года назад +4

    Now it's 2pm, I am so excited about this chip! thank you for sharing!

    • @lukasmarlon5846
      @lukasmarlon5846 3 года назад

      You probably dont give a shit but if you guys are bored like me atm you can watch pretty much all of the new movies and series on InstaFlixxer. I've been binge watching with my brother lately =)

    • @dariellian1009
      @dariellian1009 3 года назад

      @Lukas Marlon yea, have been watching on instaflixxer for since november myself :)

  • @MrTalhakamran2006
    @MrTalhakamran2006 4 года назад +1

    Useful for :
    Small to medium business sign in to office.
    Plant vs weed detection in agri.
    Detecting impurities in water tank.
    Intrusion detection.
    Tracking moving objects.
    Personalised greetings in the morning if mountes next to bathroom mirror.

  • @tinkmarshino
    @tinkmarshino 4 года назад +1

    pretty darn exciting is right! and will all the new toys coming.. it is a great time to be playing with electronics eh? Thanks for the share, carry on!

  • @avejst
    @avejst 4 года назад +10

    Great video
    Thanks for sharing :-)

  • @julianblow4739
    @julianblow4739 4 года назад +4

    :o face reveal? This thing looks so cool, I can't wait to see how makers figure out weird and innovative practical applications to this.

  • @Matt-re8bt
    @Matt-re8bt 4 года назад +3

    This would be ideal for a low-price, autonomous, search and rescue UAV. It could scour a named grid, identify subjects below, and send back properly qualified potential hits. It's a very exciting possibility.

  • @Tarbard
    @Tarbard 4 года назад +2

    Game changer. Thanks for the video.

    • @Educ8s
      @Educ8s  4 года назад +2

      I agree! I don't know why they don't market it better! I discoverd it by luck!

  • @rachidaliosinachi2065
    @rachidaliosinachi2065 4 года назад +1

    Amazing stuff !

  • @ttibor
    @ttibor 4 года назад +1

    This board is breathtaking.

  • @TheJay6621
    @TheJay6621 3 года назад +1

    I just got one.. and it is more evolved now..

  • @MrSasha3050
    @MrSasha3050 4 года назад +1

    Awesome board 😊

  • @nataliameira2283
    @nataliameira2283 3 года назад +2

    Excelente video, me ajudou muito! Obrigada e parabéns! Thanks!

  • @zetathix
    @zetathix 4 года назад +1

    Can't wait to see about maixduino.

  • @gr7480
    @gr7480 4 года назад +1

    Muy interesante. Muchas gracias !

  • @RobertShaverOfAustin
    @RobertShaverOfAustin 4 года назад +1

    I'd like to create a tripod mounted gimbal to keep a camera pointed at a person speaking on stage. It could have it's own camera or take video from the main camera. This would require the training to recognize a person's head but not necessarily the face. It would have to do the recognition for different zoom levels, that is the size of the subject would vary over some range.

  • @mghotbi4462
    @mghotbi4462 4 года назад +1

    Impressive stuff

    • @Educ8s
      @Educ8s  4 года назад

      Yes, that's trully impressive!

  • @asherasher6294
    @asherasher6294 4 года назад +1

    gonna buy this someday for face detection (target acquisition)

  • @chinmoytahbildar3278
    @chinmoytahbildar3278 4 года назад +1

    excellent... want more videos on this board... also about maxiduino... thank you

  • @viniciusnoyoutube
    @viniciusnoyoutube 4 года назад +1

    Awesome!

  • @omergecgel8903
    @omergecgel8903 4 года назад +1

    oooouuuu 😲 amazing 👏👏👏

  • @muzaffarhussainsayyed8010
    @muzaffarhussainsayyed8010 4 года назад +1

    Thanks

  • @hfe1833
    @hfe1833 4 года назад +1

    Oh man, very impressive indeed i compared the specs to openmv7 which around 60$, this is great for image recognition, im worry only about software

  • @Augmented_AI
    @Augmented_AI 3 года назад +1

    Have you tested against the OpenCV AI Kit from Luxonis?

  • @ramoncomellas4245
    @ramoncomellas4245 4 года назад +2

    Great, and I hope it will improve with the connectivity of the esp32

    • @Educ8s
      @Educ8s  4 года назад +1

      The K210 chip is the first chip the company produces! Imagine what they are going to do in a few years!

    • @hemanthkumarHere
      @hemanthkumarHere 4 года назад

      Sipeed M1W has wifi

  • @williamhart4896
    @williamhart4896 4 года назад +1

    Very interesting have heard of this board and that its fairly inexpensive as a npu to play with hadn't seen the physical product so thanks for your video .

  • @electronic7979
    @electronic7979 4 года назад +2

    Super 👍

  • @ravinduiddamalgoda7276
    @ravinduiddamalgoda7276 4 года назад +1

    Its superb 👆🔥

  • @mastermoarman
    @mastermoarman 4 года назад +1

    the maixduino would be cool if it gave you access to all the pins on the esp 32. would be perfect for robot vision machines.

  • @hemanthkumarHere
    @hemanthkumarHere 4 года назад +1

    Great .

  • @indramal
    @indramal 4 года назад +2

    I am also planning to buy this. This is best module. It is second for Jetson Nano only.

    • @Educ8s
      @Educ8s  4 года назад +1

      Yes, but it is 1/3 of the price of the jetson nano ;-)

    • @indramal
      @indramal 4 года назад

      educ8s.tv waiting for new tutorials using this device 😊

  • @supercables251
    @supercables251 3 года назад +1

    i hope this can record multiple audio channels to an sd card or something, and we can finally get sub $100 multi channel recording hardware!

  • @takeyoshix
    @takeyoshix 4 года назад +1

    Beautiful espanol accento :-)

  • @Jerseylance1
    @Jerseylance1 4 года назад +2

    I would like to use this for my humanoid robot I built for vision.

  • @benbirdsinging3275
    @benbirdsinging3275 4 года назад +1

    As you stated before, you have an accent. I don't mind, I have also an accent! English is not our own language. But do you want to lower the speaking pace? That makes everything a little clearer. Thanks in advance. The videos are educational! Thanks for that.

  • @piconano
    @piconano 3 года назад +1

    That's just scary!
    Skynet coming soon...

  • @zyroxiot9417
    @zyroxiot9417 2 года назад +2

    Greate.

  • @jparky1972
    @jparky1972 4 года назад +1

    Cool!
    My very first thoughts when I saw this?
    DIY Ironman HUD display.
    I am building my son an Ironman mask. It runs two oleds displays, an mp3 player mpu6050 and temperature sensor. Powered by a pro mini arduino.
    Looks great but very basic.
    But imagine this.
    Projected display.
    Face and object recognition with augmented reality.
    Plug that.. into Alexa and you have a very basic Jarvis!!
    Someone else can build the suit though.
    Lol

  • @MehdiGuizani
    @MehdiGuizani 8 месяцев назад +1

    😅 it is very good indeed ❤

  • @michalorganisciak9214
    @michalorganisciak9214 4 года назад +2

    Hi - I need to use external 2.2 inch LCD (or smaller) on 1,5 m cable - can I connect LCD in another way than just using this tape ?

  • @physnoct
    @physnoct 4 года назад +1

    Now, the HAL9000 can be made for cheap!

  • @rlb5261
    @rlb5261 3 года назад +1

    Excellent. I have one question, is there any way of connecting the Maixduino to the L298n+ Dc motors. The pins are identical but the voltages not. Do you have plans to do some similar like you did with the Openmv? Thank again for sharing such quality!

  • @rlb5261
    @rlb5261 3 года назад +2

    I have one question related to the Maixduino, it is similar to the Arduino except the AI capabilities that are over the board. I think about the possibility of building a robot car line follower but has a network trained to identify traffic signals or other behavior . Could you see feasible a project like that?

  • @thevivekmathema
    @thevivekmathema 4 года назад +1

    hello. can you teach how to make a custom .kmodel for your own image dataset using Keras and Maxipy IDE

  • @qzorn4440
    @qzorn4440 3 года назад +1

    sweet... the M1 or Maixduino AI Development Board looks like a fun project... how about taking an old riding lawn mower and making it autonomous or chase the cat around? Mmmm Thanks...:)

  • @fschmidtk
    @fschmidtk 4 года назад +1

    Nice video, I have this board with built-in wifi and I'm trying to connect to wifi on desktop to see the device images by IDE, would you have any help on the code to do this?

  • @aabb-zz9uw
    @aabb-zz9uw 4 года назад +1

    How can I write a program into a maixpy board? I bought one some time ago but just displays the camera image and even that doesn't work reliably.

  • @user-de9mx3kc8t
    @user-de9mx3kc8t 4 года назад +2

    Nice video! Can you use AI for networking improvements? And if you can, what parameters could you improve ?( I'm talking about the board with the esp32)

    • @Educ8s
      @Educ8s  4 года назад +1

      This board has a convolutional neural network accelerator which means it is suited for images. Sure, it can run other types of models but it is going to be slow.

  • @anirudh513
    @anirudh513 4 года назад +1

    what software did you use

  • @HarshadJoshi
    @HarshadJoshi 3 года назад +1

    can you please share your image models also?

  • @jonathanokyere-tuffuor5990
    @jonathanokyere-tuffuor5990 4 года назад +1

    Nice one. I bought one and I'm trying to use it but the when I try to download
    Maixpy IDE and it is asking for username and password

  • @josephdscordio2862
    @josephdscordio2862 2 года назад +1

    So what do usage whole of that ??

  • @RTROME
    @RTROME 4 года назад +1

    I am wondering if it can detect a certain face ( or faces programmed), to trigger an output like a relay to open a lock ?

    • @Educ8s
      @Educ8s  4 года назад +1

      Yes, it can do that after training a model to detect the faces you want

  • @berkcan3475
    @berkcan3475 2 года назад +1

    so i wonder can e connect this to pc or raspberry and run function when face is detected or something?

  • @thevivekmathema
    @thevivekmathema 4 года назад +1

    how to program or run promram on it

  • @nathanas64
    @nathanas64 4 года назад +1

    Thank you my Greek brother !

  • @nazmirnazmir1342
    @nazmirnazmir1342 3 года назад +1

    hi, i have a few questions regarding this tutorial. will u be available to answer?

  • @webslinger2011
    @webslinger2011 4 года назад +4

    I'd like to see it triggers something when the desired object is detected.

    • @Educ8s
      @Educ8s  4 года назад +4

      I can easily build a demo to trigger something when a face is detected. To detect other things, it needs more work.

    • @webslinger2011
      @webslinger2011 4 года назад +1

      Nice! Been wondering about pixy2. Haven't seen any development on it. Is there software for using rpi?

    • @Educ8s
      @Educ8s  4 года назад +3

      Pixy2 can do object detection using color quite easily. I am working on the Raspberry Pi right now. If we use tensorflow lite we can achieve object detection at about 2-3 fps.

    • @webslinger2011
      @webslinger2011 4 года назад +1

      @@Educ8s Wow! can't wait to see it. I'm currently making a wireless mobile robot car with an arm for my school project. It's in the final testing phase. I've added motioneye for field-of-view. But maybe i'll shift to an esp32 with cam since it's more compact.

  • @browaruspierogus2182
    @browaruspierogus2182 4 года назад +1

    This chip is a killer. M1w model offers fast WiFi. Risc-V 64bit cpu can be overclocked to 800Mhz so can ram.
    The only thing is the price - it should be much lower

  • @emrekeremozyurt296
    @emrekeremozyurt296 3 года назад +2

    Still esp32 looks solid and dirt cheap.

  • @gauninha12
    @gauninha12 4 года назад +1

    Does anyone know why i can't acess the bbs.sipeed.com ?? I'm trying to follow the discussions there about sipeed maix go but there is always an error when loading the page. Is it Offline? I've tried with some proxy but it didn't work. I'm from Brazil. Please someone help me :S

  • @1yyymmmddd
    @1yyymmmddd 3 года назад +1

    MaixPY IDE link is not working. :(

  • @shantosutradhar1365
    @shantosutradhar1365 4 года назад +1

    object recognition video plz plz, like pixy2 video

  • @spanchagn
    @spanchagn 3 года назад +3

    And I pay 10$ per month to ring and it can't even detect faces properly

  • @tanha8178
    @tanha8178 3 года назад +1

    400mhz, do you mean it more powerful than the pentium 350mhz that cost me 1000 buck 22 year ago?

  • @zyroxiot9417
    @zyroxiot9417 4 года назад +2

    hi, great video. Please, I want to recognize my own objects. Can you help me?

    • @darieroman10
      @darieroman10 4 года назад +3

      Create and train your own neural network with keras on your pc. Once it's done, convert to tensorflow lite, then kmodel. Flash onto the board with their utility.

  • @weithiamneo1442
    @weithiamneo1442 4 года назад +1

    Well done video! Machine learning seems interesting but always so far removed from most day to day engineering job. Could you do some tutorial how to develop AI software to run on this board?

    • @Educ8s
      @Educ8s  4 года назад +2

      Yes I will prepare a series of videos on that! Cheers!

    • @weithiamneo1442
      @weithiamneo1442 4 года назад +1

      @@Educ8s There is a similar Sipeed part number with M1W instead of M1. It seems to have built-in WIFI. Do you know if the Arduino environment is ready for this version?

    • @Educ8s
      @Educ8s  4 года назад +1

      I don't know either. The documentation is very lacking in those chips so I would stick with the ESP32 for wifi.

  • @kabirmiah75
    @kabirmiah75 4 года назад +1

    Could this be programmed to recognise individual faces?

    • @Educ8s
      @Educ8s  4 года назад +2

      Yes, it can do that after training.

  • @AhCup
    @AhCup 4 года назад +1

    Can this board do license plate recognition from it's camera?

    • @Educ8s
      @Educ8s  4 года назад +1

      Sure, if you train it correctly

    • @flyjodel
      @flyjodel 4 года назад +3

      @@Educ8s what do you mean by training ? Can you please tell us more on this process , with different application, like specific faces, plates, objets, etc ?

    • @zyroxiot9417
      @zyroxiot9417 4 года назад +2

      @@flyjodel good idea, I want too.

  • @andrewtoogood1429
    @andrewtoogood1429 4 года назад +1

    Do you have to use Micro Python. Deal breaker for me.

  • @joeybandaranaikegao4181
    @joeybandaranaikegao4181 4 года назад +3

    This proves that China has begun to be able to innovate and develop valuable electronic devices, and our price is still very friendly to electronic lovers all over the world. There is no need to discredit Chinese companies that are willing to provide convenience to the world for political reasons.

  • @balkandivision3827
    @balkandivision3827 4 года назад +1

    i would build a drone rocket to search police uniforms, and ATTACK from the sky ;D