REALLY GOOD VIDEO!!! thanks so much!!! I'm a student from HK and learning ML with your video. Your serial of Yolov8 empower me to start my AI detection project so so much. I love your patient presentation and clear pronunciation which allow me to ignore the auto caption provided from RU-vid. It's so exciting because of your high quality videos. To be honest, my blood is boiling while watching your previous video Simple YOLOv8 Class and this video! Your videos desert more view. Keep on!
Thank you so much Cheung! Ur words mean a ton to me. Happy that i Can make you so pumped haha. If u are interested I also have some courses about yolov7 for custom object detection and custom training etc. and also object tracking course with Yolov8
This is a very nice experiment that demonstrates how YOLO works as a signal for controlling the lights. You might get more fun if you obtain a custom model by going thru the transfer learning on the existing YOLOv8 so that you can differentiate the HUE of the lights upon people's preference. On the other hand, I might prefer to use the MobileNet SSD to do so since it is a simple and small model that should work well too.
An amazing project! Thank you very much. I only have one question, which may be off-topic: How do you stream the video from your cell phone to your computer? Thanks a lot!
Thanks for watching! Definitely a cool project and works nicely. I’m just using a webcam app called iriun. A bit sad that it lags so much even for the Wi-Fi connection is fairly good
@@NicolaiAI thanks could i apply some filter ( I make the tiny hole visible by passing the sheet through a dark chamber , light is passed through holes and images are taken by webcame)
Hi bro, we are working on an autonomous boat project in which we want object tracking (buoy) and the boat should move between the buoy by finding the center point between two buoys. Can you help us to achieve this so that it will be very useful for us and we could also make a promotion on our Official page about yourself?
Hey I see you have started to use a Mac for all of your most recent work. Have you given up on using your PC for all development? I am considering giving up my PC completely and maybe investing in a Mac studio instead of another PC Do you think it is necessary to still have a PC with CUDA capabilities or Mac's newest chips be suffice to handle most of our daily engineering work?
Yeah i actually just bought a MacBook Pro 16”. Mainly to be more dynamic but using it as my main computer now for coding, videos, editing, research etc since it just sync with all my devices. I still have my CUDA pc for training and gpu projects in computer vision. The Rtx 4090 blows the m2 gpu. I don’t have any work where I can utilise the gpu except editing. The gpu on Mac with mps in PyTorch is a joke for now and almost slower than the gpu. Mac + colab might be the best option for most but again main reason for a MacBook is that I can move around and work on the go
@@NicolaiAI That is completely understandable I definitely get that. So then my follow up question would be whether you would ever consider fully investing in a Mac Studio as it has the ability to out perform some of the newer GPU not necessarily the 4090 but some of the more but maybe 3*** up. To provide context I ask this question as I am about to begin my PhD in the Fall and with me going into this research heavy degree I am reluctant to invest in building a modern PC with a 4090 GPU as I feel like I am never going to use it as I stopped playing video games a long time ago now at this point and that was the main reason I got this original PC. So I am considering investing in a Mac Studio instead as I think it would better sink with all of my devices moving forward but I am nervous on whether this is the right call for me or not and wanted a second opinion from someone who has a bit more experience with all these devices.
@@keshawnsmith320 u will not be able to train anything on the Mac’s GPU. Windows or Mac is a preference. U would just use google colab pro or something for the training. I use both machines depending on the task.
