I made a pose estimation with mediapipe and OpenCV on my channel. Great framework for video processing with deep learning. Will use it more often in the future. Good video, thanks.
Hey Nicholas, first of all WOW your videos are so motivating for me and this is definitely my favourite tech/coding channel on the platform! So i'm relatively new to Machine and Deep learning and wanted to undergo a personal project where I can predict stock prices using an ML or DL model and then deploy it using an open-source tool such as flask or django. My question is: which model would you pick and why? Many thanks!
Hi thanks for your efforts and great tutorials. I guess that gray screen when you were trying to play the video is related to two things: 1. There should be a delay that actually you can see the frame. 2. The fps can solve this problem which means you need to set a 60/25 sec delay between every imshow These are only guesses, and I'm not sure I should try it by myself
One solution to gray-scale is making (writing) the three canals with the same average value from the RGB. This solution is interesting because it shows what really is a gray image.
@Geraldo, that's ONE way to get a Grey-scale image, but in normal color science, it is traditionally accepted that the "best" formula is R*0.299 + G*0.587 + B*0.114. This is due to Green being a very bright color (thus contributing the MOST to the grey-scale image) and Blue being the darkest.
Well, after having it all set up and writing the output it didnt manage to play it. When I searched for solutions, thinking I might have to use a different codes, I stumbled upon a topic on SOF saying "But there are more codecs available for mp4. You can see the list of them by setting fourcc = -1, it will show a list like this:" So I changed the fourcc argument like: "videoWriter = cv.VideoWriter(os.path.join('OpenCV', 'OpenCV_Data', 'OpenCV_Videos', 'output.mp4'), -1, fps, (width, height))" expecting a list and instead got a working video as output. It works as well, if I use 1 instead of -1. If I switch back to 'P','I','M','1' or *'XVID' or such it doesnt work. No idea whats going on there honestly.. but anyways, great videos man! Im really enjoying them and they are really helpful.
What about 3D face recognition?? And how to convert two cameras in to stereo camera?? For creating 3d face objects And Distance measurement which openCV?????
hello nicholas, sorry if im out of topic. in the previous video, u made a short video about Data Augmentation, and turns out I met that topic today in my Bootcamp class. But instead of using Tensorflow, the trainer taught us by using MXNET and D2L (which I had struggle installing the module to, lol). question is: can I just learn one library, like tensorflow? because I already dove into it, since the first day learning deep learning and CV from your tutorials. Im so confused, why are there other libraries that do the same thing (like these Data Augmentation), is it okay for me to learn just Keras Tensorflow??? THANK YOU 😊🙌
I made a pose estimation with mediapipe and OpenCV on my channel.
Great framework for video processing with deep learning.
Will use it more often in the future. Good video, thanks.
Thank you very much. Such a wonderful informative Video. Your explanations were too good. Anyone could understand.
Got the notification
And here I am
Always excited to watch your videos
THANK YOU SO MUCH. I had a problem with the codec i think, and i tried it with SO many. Didn't know PIM1 even existed TT
Bro, you are awesome! Keep making these great videos!
You are doing great man. Such hard work and amazing walk through .awesome man 👍 keep them coming.
See ya. Going to explore your channel now.
Very nice explanation with suitable example. Thanks
Thanks for such an Awsome Tutorial, I am going to mix it up with your mediapipe tutorial.
Great stuff, subbed on the spot!
Thanks Nicholas. Your videos are a big help to me.
Hey Nicholas, first of all WOW your videos are so motivating for me and this is definitely my favourite tech/coding channel on the platform! So i'm relatively new to Machine and Deep learning and wanted to undergo a personal project where I can predict stock prices using an ML or DL model and then deploy it using an open-source tool such as flask or django. My question is: which model would you pick and why? Many thanks!
Facebook prophet or neural prophet
Hi thanks for your efforts and great tutorials. I guess that gray screen when you were trying to play the video is related to two things:
1. There should be a delay that actually you can see the frame.
2. The fps can solve this problem which means you need to set a 60/25 sec delay between every imshow
These are only guesses, and I'm not sure I should try it by myself
One solution to gray-scale is making (writing) the three canals with the same average value from the RGB. This solution is interesting because it shows what really is a gray image.
@Geraldo, that's ONE way to get a Grey-scale image, but in normal color science, it is traditionally accepted that the "best" formula is R*0.299 + G*0.587 + B*0.114. This is due to Green being a very bright color (thus contributing the MOST to the grey-scale image) and Blue being the darkest.
Really awesome tutorial, thank you very much.
Great video. Thanks for sharing.
Well, after having it all set up and writing the output it didnt manage to play it.
When I searched for solutions, thinking I might have to use a different codes, I stumbled upon a topic on SOF saying
"But there are more codecs available for mp4. You can see the list of them by setting fourcc = -1, it will show a list like this:"
So I changed the fourcc argument like:
"videoWriter = cv.VideoWriter(os.path.join('OpenCV', 'OpenCV_Data', 'OpenCV_Videos', 'output.mp4'), -1, fps, (width, height))"
expecting a list and instead got a working video as output. It works as well, if I use 1 instead of -1. If I switch back to 'P','I','M','1' or *'XVID' or such it doesnt work.
No idea whats going on there honestly..
but anyways, great videos man! Im really enjoying them and they are really helpful.
A video on GStreamer and all its horrors would be good to do
Can u make a video on driver drowsiness using CNN architecture??
As I know the waitKey is essential for the imshow method to work. Only when waitkey is called the video is displayed.
What about 3D face recognition?? And how to convert two cameras in to stereo camera?? For creating 3d face objects And Distance measurement which openCV?????
Ooooh, haven't touched on that yet! Could possibly take a look in the future @Daniel!
hello nicholas, sorry if im out of topic. in the previous video, u made a short video about Data Augmentation, and turns out I met that topic today in my Bootcamp class. But instead of using Tensorflow, the trainer taught us by using MXNET and D2L (which I had struggle installing the module to, lol).
question is:
can I just learn one library, like tensorflow? because I already dove into it, since the first day learning deep learning and CV from your tutorials. Im so confused, why are there other libraries that do the same thing (like these Data Augmentation), is it okay for me to learn just Keras Tensorflow???
THANK YOU 😊🙌
Amazing tutorial!!
Hello sir, How to generate summary of the vedio?
I'm working on this project and stuck.. could you please help me?
Convert the video to audio first, then audio to text then finally using a summarizer!
Hey nic, can you suggest me a way of presenting visually the results of the sentiment analysis I've done
how can i make my own 3d camera tracking with python
analyzes video sequences to extract camera motion and 3D scene data
thank you 3000
Thanks
Iam getting an error while writing a video in opencv 4.7.072
which is the best codec for mp4 files?
how you use auto typing/auto correct at 15:58 in JupyterLab?
Mr:nichola
The video is running too fast as 2×
Bruh all of this is cap
(got the joke? cap? capture?)
*crickets*
nah, I got it LOL
the guy who pretend writing a code 😁