This channel is about www.rocksetta.com a website to display 40 years of computer programming ideas, from cryptocurrencies, Machine Learning, Javascript Teaching, App Making with Phonegap, DIY Makers, Hockey Training, Rugby Learning, 3D Printing, Robotics and Kickstarter inventing along with my learning Music ideas. It was made to help people who have given up on being able to play an instrument or sing a song. Rocksetta tries to make music more visual. Combining rocksetta with keyfreemusic may help some people learn to sing or play. This channel is also about my experiences teaching Robotics and 3D Printing animation and computer coding.
Hello. The Maker100=Eco Rak2270 w/Machine Learning is extreme interesting, is there somewhere I can buy this setup, or commission someone to make these. I doing a project for elderly people with cognitive diseases.
@Blumainor which one there are three types. store.rakwireless.com/products/rak2270-rak-sticker-tracker?srsltid=AfmBOooZQt_xxXU7Yk9zmbf-JEKu5PjYjDp6gz9zF6tMWgjfk0Gr28ln&variant=43194331791558 also I am quicker to reply to a DM on linkedin ca.linkedin.com/in/jeremy-ellis-4237a9bb
Thank you for all your work on this board! I am eager to hear of your example used for passing variables between cores. Do you have a video on this that I am missing?
@@kaijustice7048 actually I prefer dot42 but yes there are lots of ways to communicate between the cores. sometimes I just give up and use threads in a single core. One cool trick if you are just turning things on or off is to use an LED. Core M7 can turn it on and core M4 can read if it is turned on. Dot42 is the best UART I have used and uses the pins that are visible from the Portenta without the HD connectors. github.com/hpssjellis/portenta-pro-community-solutions/blob/main/examples/dot4-portenta-breakout-board/dot42-UART1-M4-to-M7/dot42-UART1-M4-to-M7.ino
Yeh, it is a bit scary if your not good at soldering. I recently got a soldering microscope and it makes things like that a lot easier. Not sure why Arduino doesn't allow you to purchase them with the headers pre-attahed for a few extra bucks.
Hi Jeremy, you most likely won’t remember me. In Grade 9 I was “unwillingly” put into your Web Design class at HPSS in 2005 thinking to myself how life couldn’t get any worse. Two days into the class I fell in love with coding. I could get the internet to do whatever I want! I even built multiple minor hockey websites while in your class and till I graduated high school… I made great use of inline CSS 😂 Now here I am today 18 years later and still coding. I have greatly expanded my knowledge in all aspects of web development and have built many upon many websites, databases etc… apart from web development I have launched 2 apps into the Apple App Store, built some pretty neat projects with the Arduino, explored into game dev with Unity and Unreal Engine and VR, taught myself VBA, AHK, Python, Objective C, Batch and more. I was a lead programmer for Best Buy Canada writing all the shipping, receiving and quality control software. Now I currently run my own computer repair business in Mission and develop/code on the side while also working for the City. Coding is a huge part of my life and I would have never been where I am now if it wasn’t for you. I often think back to the first day I was In your class where it all began. I hope all is well during these tough times and 18 years later I would like to sincerely thank you for being the model and influence that made me the developer I am today. -Steven Mayer PS- It you don’t remember me, I was the kid that thew another kid over the computer desk for licking my ear and you definetly woke the school up yelling at us 😅
Thank you so much for this video. I was so stuck I almost bought another laptop. Been trying to run linux on my laptop since December 2022 and boom I did it.
The best video I've seen all month. Thank you for bringing this to my attention. Certainly sharing with the team internally. The passion is next level. Thank you for being part of the wonderful community of web ML!
Ahem. First off:wow. Your body of work is so impressive on its own. I have been fumbling about in the weeds with Portemta for a full week, I sure wish I'd bumbled here sooner! I am considering Opta for some (rail comms) integration, and I was getting the vapors. But you have given me awesome hope. Thanks!
Hello sir, I saw your work on portenta h7, I am reapply inspired by your work. i am working on computer vision for my final year project, using portenta h7 and portenta vision shield. i have completed most of the work but I am getting some errors with figuring out hardware problems, I would be really thankful to you if you will guide me on my project.
@Rocksetta I am making a real time facial recognition, I trained my model from edge impulse and deployed it using openmv, now the problem is that I connected my portenta h7 and portenta vision with portenta breakout, and I want to connect my spi, tft lcd to make it stand alone but I can't figure out pins for connecting the lcd from portenta breakout.
@@SOFIA_TECH_KITCHEN Yes a TFT has an issue connecting with the PortentaH7 see this link on my course github.com/hpssjellis/maker100#29 I much prefer the waveShare 128x128 grayscale OLED, cheap like $20 and it is much faster than the typical TFT. Here is my link to working with the grayscale OLED github.com/hpssjellis/maker100#24 Good luck. That is tricky stuff. I never did get the touchscreen part of the TFT working well. Do you need touchscreen or can you control the program some other way. By the way facial recognition is really hard on these tiny devices, lots of luck to you with your model.
Thank you very much for your videos. I am working with Portenta h7, and I want to connect it with the classic Bluetooth to my phone and control my toy car with MIT. Do you have any code example of classic Bluetooth?
@Jeremy Ellis i have been working on my project for some days now. I am a student online. But don't know how to check my project on web . I am doing my project on gitpod. So could you explain how i could see what progress i have made so far please
@@brany2920 Gitpod has changed a bit since I made those videos. I need to know more about the issue. You are welcome to follow me on Twitter @rocksetta and I we could DM possible ways to help, but for now can you clarify what works? Can you see a webpage just not from outside the work environment? If that is the case there is a globe icon that allows your gitpod to be public. Having said that gitpods are only active while you are logged in, when you leave they shutdown. Reply a bit more about the issue.
That is why I switched to the Arduino PortentaH7 with LoRa vision Shield. It is just a bit more powerful and has 8 MB SDRAM which helps with using a Greyscale OLED.
@@Rocksetta a raspberry pi 4 model b can run a tflite object detector (efficientDetLite2, to be specific) at about 4 inferences a second. Add a CORAL Edge TPU to it, it shoots up to a staggering 14 frames/inferences a second.
congratulations for the video. I would like your help to turn on an RGB led of the nano board with the gestures I saved as a sample. what can i do, friend?
congratulations for the video. I would like your help to turn on an RGB led of the nano board with the gestures I saved as a sample. what can i do, friend?
congratulations for the video. I would like your help to turn on an RGB led of the nano board with the gestures I saved as a sample. what can i do, friend?
I am still pondering how to merg face armatures with the plask downloaded objects. I will make another video if I figure it out. Great product plask.ai
Great job mate, would you mind giving us a bit more info on the dataset you used? I'm only asking because I have my eye on doing a similar project, but all the images in that dataset are 200 x 200 pixels by default and I'd need to try and adapt it (in Edge Impulse) for the OV7670 module.
Hey Martin, the nano33BleSense is a great board but for vision work I have moved to the portenta with GrayScale vision shield. Since I use the Pixy2 from charmedLabs for all color stuff. As far as the data set, I make my own. Supper easy with a cell phone on edgeImpulse. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-wbX_-No8kIw.html
Great Video. :) Short addition, the training result is always different because the weights are initialized randomly each time. The learning rate is set relatively high in the examples, so it may be that no global minimum is found. Each training is individual and it will not be trained further if the command is entered again.
Yes, any Arduino or Arduino compatible board. AIfES even runs on the PC.Tutorials are available on our Arduino Project Hub page: create.arduino.cc/projecthub/aifes_team
My one was very light, which is not so good as any contact makes it move. A heavier one might be a good idea. Your idea makes sense. Send a picture to twitter @rocksetta if you ever get it made.
Is it possible to run micropython and arduino (C++) code simultaneously? Or does OpenMV uses both cores? Btw you are very helpful here and on the forums
Marian, last I checked about 6 months ago, OpenMV takes all the memory for both cores. So no, can't be done yet. Things may have changed, please reply if you find out. The openMv forum is friendly. I would definitely do a video on it. New stuff in my portenta library.
Nice to see this working with the Arduino IDE. It's a nice board. I saw this on Mouser.ca with the following part numbers: Mouser Part # 246-CXD5602PWBMAIN1E Mouser Part # 246-CXD5602PWBEXT1E Mouser Part # 246-CXD5602PWBCAM1 Are you able to confirm the part numbers of the Sony Spresense boards you have? I'd like to try them out with Arduino and Edge Impulse for ML inference. Thanks so much Jeremy! :)
Hi Mark: The numbers on the boxes I have are very similar to what you have quoted: CXD5602PWBMAIN, CXD5602PWBEXT1, CXD5602PWBCAM. I have not fully got mine working with Edge Impulse yet, but am so far reasonably impressed with the compile speed on the Arduino 1.8? IDE. Not sure how much work I will do with the Spresense as my main teaching Robotics and Machine Learning interest for 2022 is the Arduino Portenta with LoRa and Ethernet Vision Shields and Breakout board.
@@Rocksetta Thanks for taking the time to confirm the part numbers for me, Jeremy. It's cool to see Sony offering a suite of hardware and accessories for the embedded IoT space, especially with future proofing TinyML inference with the ARM Cortex M4F based MCU. I like the fact it also has built in GPS in a small package at a reasonable price. Hope you're able to get it up and running on the Arduino IDE very soon. I'm going to check out the developers setup page to see how to do this. It's a good board and hardware platform like the Arduino Nano 33 BLE Sense. Keep up the good work! 👍 I've been studying MEMS audio sensor information over the past week for a project related to TinyML on an embedded system. A MEMS microphone array is the way to go for digital audio detection and direction finding of various human generated and human labeled audio sound sources. I'm looking into developing a solution for terrestrial and aerial based applications for a pet project that I can use as a demo for prospects and clients. I'll share this on Hackster.io as I get a little farther ahead.
Hey Mark, please make an account on the Edge Impulse forum forum.edgeimpulse.com/ and tag me @rocksetta That pet and microphone array sounds really interesting. The people at Edge Impulse would be very interested and they also have great coding ideas. Tag @jangongboom as he is brilliant and codes what I take a month to code in minutes.
