Mind blowing tutorial! Look forward for more similar projects, especially the one you talked about (counter for several exercises at once). Liked + subscribed + bell turned on!
Hi Mr. (@nicholasrenotte) Ronetta, you are giving great examples. In this study, I tried to measure the shoulder, knee and ankle angles of a person shooting a basket from the side, but I could not. Can you share with us an example perceived from the side? Even if you don't have the time, thank you very much, even these studies increased my ability.👏👏👏
@@sobanrauf7649 yes model deployment onto cloud is possible using AWS, Azure or GCP with docker. Or another way is to build the whole app using the model and then deploy the app using the above mentioned ones. I would prefer way 2, now since I don't know Html, css or react, I make web apps with streamlit and other options and deploy them.
@@dipankarnandi7708 will it work same when uploaded to model and embeded to app as it is working here with open cv would i have to change tha code in model or upload same?
@@sobanrauf7649 you gotta use the cloud features here. Do this, first create the app and embed the model in it. See if it works, If it works there, then all you gotta do in containerize it with Docker / AWS Sagemaker to deploy wherever you want. Go step wise
Thank you dude! You are the most underrated youtuber I have ever seen I'm loving this video although I have one question How do you do the exact same thing But instead of a camera, You do it with screen recording live? Hope you or anyone else can help me Cuz I'm really stuck
Could access a screen using pyautogui and use that as the video instead of the feed from opencv. Check this out www.thepythoncode.com/article/make-screen-recorder-python
Thank you so much. Your teaching and tutorial style is the perfect match for me. Clear an precise, and you continuously explain everything you do, thanx!!
Thank you sooo much in trying to help, but could you please copy and paste your reply about my question again to me? for some reason you tube hid that reply and I cannot get it. Please do so as it is important to me. Thank you again.
Hello Sir, my name is rizky and i want to ask about how to removing landmarks and only showing landmarks we needed (for example if i only want hand landmark pose just show the hand only) ? Let me know if you have a good advice for that and your tutorial is awsome 🙏
"Make Detections" and the parts after that doesn't work. I run the codes but nothing happens. Do you have any idea about that? Could you please help me? Thank you so much.
Dude you are awesome. !!! Did the multi joint estimation video out, if its there cant find it out !!! I wish to make a yoga pose detection... Thanks in advance
Great Video Nicholas! Thank you very much. I have just one question- I am having a problem installing mediapipe. I used "pip install mediapipe" but it didn't recognize the library. Would appreciate the help. It's kinda urgent.😄
I'm having an issue, when i ran the code just in beginning camera turns on I got black in white screen and than Unknown C++ exception from OpenCV code appears on jupyternotebook. Is this some hardware problem or what? please help!
I am getting this error when i use pose = mp_pose.Pose(). : SplitLandmarkListCalculator: The number of output streams should match the number of ranges specified in the CalculatorOptions. No sollution available on the Internet. Can you help?
Can you explain more about the formula arctan in this video. Are there any sources for me to learn about that . it took me about 2 hours to finish this projects but I do not really understand how the formula works. Thanks!
Check this out: www.mathopenref.com/arctan.html#:~:text=The%20arctan%20function%20is%20the%20inverse%20of%20the%20tangent%20function.&text=Means%3A%20The%20angle%20whose%20tangent,to%20know%20the%20actual%20angle.
Your all tutorials are so amazing and productive. In this lesson, I want to just display the angle in a whole number, not a integer. How can we do this. By the way thankyou for creating a nice tutorial
I did everything right . No error is shown but the results are not coming for data collection. Why ? It says that it can't read 'q' in the code . If anyone knows what it means , just let me know
Hi Nicholas, first of all thanks. I tried to open the webcam feed as you told in Jupyter notebook, not able to get any feed bro ..not even any errors ? I also changed the id from 0 -5 but things didn't work out. Tried the same with GitHub and python platforms No help. Can you please let me know what can be the possible source of error for this to happen ? Thanks in advance !!
HELLO friend...i love what you did here...so awesome...good job...do you think if i dont know anything of python I can do it myself? capturing, I mean...can I install and run this on windows?? but what platform do i use? (run in where?) im sorry if i am confusing you...
@Nicholas Renotte Hai , how to calculate time duration between two human action or human poses , for example how to calculate hand wash time duration of a human in video any suggestion and references are most welcome
Im not a coder. Im in healthcare and i would really like to try this for a patient of mine. How could I get this going? i know I have dependencies to install....is there a basic video tutorial?
Heya @W Gallo, first up you would need to install Python (www.python.org/downloads/) and Jupyter Notebooks, just run: pip install jupyter notebook at a command prompt to install it. Then from there you can download and use the code I've got in the description :) let me know if you get stuck, I'm happy to help!
Hey Nick, just wondering when does programming end and data science begin? I've spent nearly a year learning programming and math. Pure programming. Syntax et al Python. I've just finished a chapter on classes and modules. Do data scientists rely on classes and modules a high proportion of the time? I don't want to get stuck in programming too long. Thank you in advance.
As much as I'd like to say there's a clear cut line, there really isn't @Oscar. The best data scientists I've met tend to be highly adept at software engineering as well as data science principle, this is doubly true for deep learning and machine learning engineers. That doesn't mean you should hold off on learning DS now though. You're going to continually need to learn new things as your experience grows. My advice, start data science now! As soon as you've got a basic understanding of programming start working with data and data science libraries ASAP. I took me a long while to jump from regular coding when I was a consultant to working with machine learning because I thought I needed a ridiculously strong base. What I found was that I was I was more than capable enough with basic Python principles. Occasionally I found I had to double back to remember how to do some of the more complicated stuff, building classes and setting up class inheritance for example, but normally it's not hard enough to stop me progressing with the current task.
Bro nice❤ video, can plz help me i want to built a app that can detect pose or pose estimation for exercise count.. Can you tell me from where i can start what i need to built that type of android app
hey nicholas.. how can we create a 3d map of surrounding using opencv as it can serve applications like calculating distance between joints and many other for developing an AI trainer
As a beginner coder, I found this tutorial helpful and very easy to understand. There is only one part which I can not seem to grasp. I have gotten the code correct and even tried copy and pasting the one in the GitHub, but I am unable to get the angle to display on my screen. My assumption is that I have gotten the webcam dimensions incorrect (which you put as 640, 480), and I am not sure how to find out what the correct number should be.
Hello There, I have a question related to angle measurement, I want to get the angle for bowling action analysis... For that I need to calculate an angle between Shoulder, elbow and hand wrist... Form this video the angle is continuously measured from the camera but I want to calculate angle only when user perform a bowling action. For the ICC rule, the angle is measured such that first we calculate angle once when user arm is parallel to ground (angle1) and when he throughs the ball (angle2) , after that I need to subtract the angle2 from angle1.... Please guide Thanks
Hi Nicholas! First of all thanks for amazing tutorial. It was really helpful. I love your tutorials so much you are explaining the import details in very understanding way. But I want to ask you a question. I did not really fully understand the angle calculation part and I could not find anything about it on the internet either. Could you please give a little bit more info about that part?
Can you please give me a little help?? question 1: If i am just plying a video of a person and i want to visualize the coordinates next to the respective joint, what is the code i use? could you write it here? with the webcam you put cv2.puttext (image, str(angle), tuple...i guess my doubt concerns the multiplication of 640, 480.. question 2: if i want multiple joints together do i just keep listing the #get coordinates without separating anything? in other words i would want both hips, both shoulders, both elbows, knees, etc. question 3: how do i do if i want to put a true/false to turn off just the display of the angles? (im not a coder but i was able to get it all running because of your teaching, so thank you a billion!!! Good job on it!!)
Hello Nicholas, Great work doing all of these; it was enlightening. I wanted to know if you have done the multi-pose/ multi gym tracking you spoke of around 52:11. Thank you.
@@NicholasRenotte Thank you so much! Is there a way to estimate if specific key points in a video fed into the MediaPipe are missing? Say a video of me with just my head and shoulders, and I want to calculate the key point of my ankle. Is this possible?
Appreciative for your effort to teach me a comprehensive about python especially mediapipe this tool help me too much 😜😜 But nick what about hand tracking and caculate the angle between wrist and thumb, i have been tried however it not work 🥲
@Nicholas Renotte while running make detection code, camera window pops up but after pressing "q" button to close camera window my notebook's kernel shows error that kernel is died and it gets restart and then I can't go further because of kernel died again and again. What's the issue.?
@@NicholasRenotte sorry for not keep track of my notifications. the issue was resolved by making some version controls for some libraries, now my question is, if i want this real-time application to show me the text"XYZ" when right_shoulder, right_hip and right_knee make 90-degree angles?
I love it very much! Is it possible to calculate how accurate our poses are in order to make sure that our poses are correct comparing with gym trainer ? Thanks!!
i am getting this error even though i have install all dependencies correctly D:\PrOjEcT\PoseEstimation>python PoseEstimation.py Traceback (most recent call last): File "PoseEstimation.py", line 12, in cv2.imshow('Mediapipe Feed', frame) cv2.error: OpenCV(4.5.1) C:\Users\appveyor\AppData\Local\Temp\1\pip-req-build-1drr4hl0\opencv\modules\highgui\src\window.cpp:651: error: (-2:Unspecified error) The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Cocoa support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script in function 'cvShowImage'
When I tyoe the part that includes 'with mp_pose.Pose(min_detection_confidence=1, min_tracking_confidence=1) as pose:' I get this error, 'RuntimeError: ValidatedGraphConfig Initialization failed', is this project able to be completed on the CPU of a laptop and if so do you know of any solutions to my error, i have looked on google and couldnt fins a solution
Hey Nicholas, this is my first time using mediapipe. Why am I not getting the medipipe feed when I "print(results) in the "Make Detection section of the code (i am using Google Colab)?
How does camera plays a role in Open-pose detection. Because my camera (laptop-webcam) is of low Mega-pixel, what if i use smartphone camera instead. How does it affect the pose-estimation. Thank you
Is there any way to get a somewhat workable hand tracking working on a pi zero or even a microcontroller like a esp32? Im struggling to find any real approaches and starting to think that there are none, but i dont want to give up yet. Mediapipe on the pi is not running, even after adjusting the os version and installing hundreds of apparently missing packages Installing the python version on an esp32 is not possible since thers only micropython and it doenst support these high-level-packages. Since many examples for the esp are written in c++ i had hoped to be able to use the c++ version of mediapipe, but installing this is quite the opposite of straightforward and i dont even know where to start. Every project i found with has hand tracking and a esp32 in it, makes use of the .js version where the calcutation is outsourced to an client Thats not an option for me Am I chasing ghosts here? Is that simply not possible? Even with very few frames?
Thanks for your great tutorials. Is there any tutorial, that shows how we can customize/modify or retrain the Mediapipe models? Current models are trained for adult pose detection and have issues in pose detection in infants. How can I improve it?
Hey friend...this is AMAZING!!!! Can you help me with a question??? im following your steps but cannot go beyond step 2. After telling me it successfully installed mediapipe, after the first step, it tells me that it cannot recognize this line: "import mediapipe as mp" ?? what am i doing wrong?? could you please enlighten me?? Thank you for any help!!!
Hello when trying to use mediapipe in jupyter lab I run the error "module 'mediapipe' has no attribute 'solutions' " Do you have any clue how to fix this?
YES YES YES!!! The video I more than needed. So helpful! Thank you!!! If i wanted to use this method to calculate work done (WD= Force x Distance), how would you suggest I go about it. My thinking is that I'd assign a certain weight to either the joints or the line between joints and use that for force (signifying weight distribution in the body) then the model tracks the distance each joint or line moves throught space and multiplies to get work done. The individual work done per joint or line is summed and I'd have total work done by the body. Practicalizing this is what I can't seem to figure out.
I actually looked into this a while ago, it'd probably be something along the lines of angular force. Take an assumption of the mass of the arm, then calculate it's accleration (speed between frames) then assuming you know th length of the arm or use a proxy e.g. coordinate to coordinate euclidian distance...you should be able to do it!
@Nicolas Thanks it is such great help. I have 2 quick questions >>>> why have you multiplied in angle code np.multiply(elbow, [640, 480]).astype(int) >>>> how to resize the video frame if you can suggest some learning sources on same. It will be wonderful. Big thanks again man
Heya @Karamvir, 1. This is because we receive normalized landmarks we need to return them to our frame size to render them based on our baseline image 2. Can resize it by adding cv2.resize before the render :)!
Like, I don't understand why in this formula you do not use the index-2 coordinates of the points: radians = np.arctan2(c[1]-b[1], c[0]-b[0]) - np.arctan2(a[1]-b[1], a[0]-b[0])
Visuale angle and counter logic not working write the code with PyCharm, could you please help :) if you send me .py version I would be very glad to you. Btw I use Mac but I'm going to make the project to use Jetson
@@NicholasRenotte Hi, I solved it, totally my fault :) I was forgot the "def" :D so it could not calculate the angle :) Thank you so much, the code is very useful for our project. By the way, our project is about the physical activity at the bus stops, with this code we will guide the people during waiting. If they complete the movements, system will donate one turkish lira to ALS foundation. Thank you again, be safe and take care.
I copied your code, but the angle is not displayed on the screen. I don't know what the problem is.
if you found a solution please tell me
Did you find a solution? I'm dealing with the same problem
@@kaviyak7308 did you find a solution?
Sir, your presentation is really systematic and content is as always informative. Thank you so much.
Thank you so much @Winston, glad you enjoyed it!
Mind blowing tutorial! Look forward for more similar projects, especially the one you talked about (counter for several exercises at once). Liked + subscribed + bell turned on!
When will that video come out
Hi Mr. (@nicholasrenotte) Ronetta, you are giving great examples. In this study, I tried to measure the shoulder, knee and ankle angles of a person shooting a basket from the side, but I could not. Can you share with us an example perceived from the side? Even if you don't have the time, thank you very much, even these studies increased my ability.👏👏👏
I think he mentioned that this method has limited potential particularly for a single joint. For multipose, try using other method
@@dipankarnandi7708 can u please guide if i can upload same model to cloud and use it in my android react native app
@@sobanrauf7649 yes model deployment onto cloud is possible using AWS, Azure or GCP with docker.
Or another way is to build the whole app using the model and then deploy the app using the above mentioned ones.
I would prefer way 2, now since I don't know Html, css or react, I make web apps with streamlit and other options and deploy them.
@@dipankarnandi7708 will it work same when uploaded to model and embeded to app as it is working here with open cv would i have to change tha code in model or upload same?
@@sobanrauf7649 you gotta use the cloud features here.
Do this, first create the app and embed the model in it. See if it works,
If it works there, then all you gotta do in containerize it with Docker / AWS Sagemaker to deploy wherever you want.
Go step wise
Thank you dude!
You are the most underrated youtuber I have ever seen
I'm loving this video although I have one question
How do you do the exact same thing
But instead of a camera,
You do it with screen recording live?
Hope you or anyone else can help me
Cuz I'm really stuck
Could access a screen using pyautogui and use that as the video instead of the feed from opencv. Check this out www.thepythoncode.com/article/make-screen-recorder-python
@@NicholasRenotte thanks!
After MediaPipe Please do cover YOLO v5 /v4 ! I love your courses. They are really awesome! :)
Namaste sir
😂😂😂❤️❤️
Ayyyeee, you got it!
Thank you so much. Your teaching and tutorial style is the perfect match for me. Clear an precise, and you continuously explain everything you do, thanx!!
Thank you sooo much in trying to help, but could you please copy and paste your reply about my question again to me? for some reason you tube hid that reply and I cannot get it. Please do so as it is important to me. Thank you again.
My bad, what was the question again? Just catching up on comments now.
here for the youtube algorithms. I will definitely watch it later tho. :)
Thanks so much @Mike, let me know what you think!
mythmon spotted :3
Hi,
Can we use a depth camera and put real depth value (z value ) instead of its estimation as (result.pose_landmark.z)???
Hello Sir, my name is rizky and i want to ask about how to removing landmarks and only showing landmarks we needed (for example if i only want hand landmark pose just show the hand only) ?
Let me know if you have a good advice for that and your tutorial is awsome 🙏
Easiest to do it by just using cv2.circle to draw the landmarks that you want from results.pose_landmarks
From result.post_landmark
Should We took the coordinates from landmark in bodypose then use cv2.circle ?
"Make Detections" and the parts after that doesn't work. I run the codes but nothing happens. Do you have any idea about that? Could you please help me? Thank you so much.
hi , thanks for your video
i would ask to the learning source of calculating joint angels , it would help me alot thanks in advance :)
Mannnn, I can't remember where i got it from. It was from somewhere on the TF site.
THANK YOU SO MUCH. YOU SAVED MY SENIOR DESIGN PROJECT
Ayyyyy, awesome work my guy!
Dude you are awesome. !!! Did the multi joint estimation video out, if its there cant find it out !!! I wish to make a yoga pose detection... Thanks in advance
Name 'landmarks' is not defined this is my error
Great Video Nicholas! Thank you very much. I have just one question- I am having a problem installing mediapipe. I used "pip install mediapipe" but it didn't recognize the library. Would appreciate the help. It's kinda urgent.😄
Shoot, might be your Python version, can you try 3.7.3 instead
I'm having an issue, when i ran the code just in beginning camera turns on I got black in white screen and than Unknown C++ exception from OpenCV code appears on jupyternotebook. Is this some hardware problem or what? please help!
Can you shoot me the error @Waseem? Could be a bunch of things but it sounds like Visual C++ build tools might be missing.
@@NicholasRenotte I noticed an error which says "unknown c++ exception from opencv" in jupyternotebook!
@@clashcascade hmmm, can you show me the exact error?
mediapipe feed is not coming what should I do?
I am getting this error when i use pose = mp_pose.Pose(). : SplitLandmarkListCalculator: The number of output streams should match the number of ranges specified in the CalculatorOptions. No sollution available on the Internet. Can you help?
Can you explain more about the formula arctan in this video. Are there any sources for me to learn about that . it took me about 2 hours to finish this projects but I do not really understand how the formula works. Thanks!
Check this out: www.mathopenref.com/arctan.html#:~:text=The%20arctan%20function%20is%20the%20inverse%20of%20the%20tangent%20function.&text=Means%3A%20The%20angle%20whose%20tangent,to%20know%20the%20actual%20angle.
I CANT SEE ANGLES IN MY FEED......ELSE EVERYTHING IS PERFECT......CAN ENYONE EXPLAIN ME WHY?
Your all tutorials are so amazing and productive. In this lesson, I want to just display the angle in a whole number, not a integer. How can we do this. By the way thankyou for creating a nice tutorial
PLEASE HELP : I paste to script in PyCharm but nothing works, i have no errors but i can only see my face in the webcam
Hii bro
I am unable to install mediapipe....
I did everything right . No error is shown but the results are not coming for data collection. Why ? It says that it can't read 'q' in the code . If anyone knows what it means , just let me know
Very very helpful. Glad I found it. Thank you ❤.
Hi Nicholas, first of all thanks.
I tried to open the webcam feed as you told in Jupyter notebook, not able to get any feed bro ..not even any errors ? I also changed the id from 0 -5 but things didn't work out. Tried the same with GitHub and python platforms No help. Can you please let me know what can be the possible source of error for this to happen ? Thanks in advance !!
Ha weird, check this out, might help! github.com/pvys/CV-camera-finder
Cool, dude..
How we can show 2 or more coordinate angle?
Check this out: th-cam.com/video/EgjwKM3KzGU/w-d-xo.html
HELLO friend...i love what you did here...so awesome...good job...do you think if i dont know anything of python I can do it myself? capturing, I mean...can I install and run this on windows?? but what platform do i use? (run in where?) im sorry if i am confusing you...
Sure can! I've got a series on Python from scratch coming soon if it helps :)
Hii sir, i executed same code which was you executed.But i couldn't get angle also i couldn't get count also.what is the problem sir
@Nicholas Renotte Hai , how to calculate time duration between two human action or human poses , for example how to calculate hand wash time duration of a human in video any suggestion and references are most welcome
Im not a coder. Im in healthcare and i would really like to try this for a patient of mine. How could I get this going? i know I have dependencies to install....is there a basic video tutorial?
Heya @W Gallo, first up you would need to install Python (www.python.org/downloads/) and Jupyter Notebooks, just run: pip install jupyter notebook at a command prompt to install it. Then from there you can download and use the code I've got in the description :) let me know if you get stuck, I'm happy to help!
Hi, are you using which model of BlazePose (BlazePose.Heavy or BlazePose.Full or BlazePose.Lite)?
Hey Nick, just wondering when does programming end and data science begin? I've spent nearly a year learning programming and math. Pure programming. Syntax et al Python. I've just finished a chapter on classes and modules. Do data scientists rely on classes and modules a high proportion of the time? I don't want to get stuck in programming too long. Thank you in advance.
As much as I'd like to say there's a clear cut line, there really isn't @Oscar. The best data scientists I've met tend to be highly adept at software engineering as well as data science principle, this is doubly true for deep learning and machine learning engineers.
That doesn't mean you should hold off on learning DS now though. You're going to continually need to learn new things as your experience grows.
My advice, start data science now! As soon as you've got a basic understanding of programming start working with data and data science libraries ASAP. I took me a long while to jump from regular coding when I was a consultant to working with machine learning because I thought I needed a ridiculously strong base. What I found was that I was I was more than capable enough with basic Python principles. Occasionally I found I had to double back to remember how to do some of the more complicated stuff, building classes and setting up class inheritance for example, but normally it's not hard enough to stop me progressing with the current task.
Hi Sir, can you make face filter with mediapipe?
Thanks
Bro nice❤ video, can plz help me i want to built a app that can detect pose or pose estimation for exercise count.. Can you tell me from where i can start what i need to built that type of android app
how can we overcome that even if we hit on our shoulder also its taking count
Man, you're wearing some tight jeans!! But amazing video , helped a lot
what if❓ I had pre-recorded video 🎬and want to apply pose estimation 🦿🦾 on that video.
hey nicholas.. how can we create a 3d map of surrounding using opencv as it can serve applications like calculating distance between joints and many other for developing an AI trainer
Heya @Jasdeep, take a look at SLAM using Lidar for that type of use case :)
As a beginner coder, I found this tutorial helpful and very easy to understand. There is only one part which I can not seem to grasp. I have gotten the code correct and even tried copy and pasting the one in the GitHub, but I am unable to get the angle to display on my screen. My assumption is that I have gotten the webcam dimensions incorrect (which you put as 640, 480), and I am not sure how to find out what the correct number should be.
Did you find the solution for it ??
i'm unable to get any camera output with my apple m1, anyone faced similar challenge?
How can we deploy this mobile into a mobile app?
Could do it with Tensorflow JS and React?
Sir can you say what type of algorithm you follow. Like CNN, GNN.
Hey you mentioned you will you a multi exercise tracker in another video, have you done that already?
Is there a way to calculate the human neck angle using this approach?
If we want to build crunches counter what is the logic for going up and down? could someone tell
Hello There,
I have a question related to angle measurement, I want to get the angle for bowling action analysis... For that I need to calculate an angle between Shoulder, elbow and hand wrist... Form this video the angle is continuously measured from the camera but I want to calculate angle only when user perform a bowling action.
For the ICC rule, the angle is measured such that first we calculate angle once when user arm is parallel to ground (angle1) and when he throughs the ball (angle2) , after that I need to subtract the angle2 from angle1....
Please guide
Thanks
Haven't tackled that @Waseem, might need to do some research.
@@NicholasRenotte Please let me know If there's something that might help me to get this done as I have searched alot...
how to calculate angle of neck ??
I cannot see the angles in my feed otherwise everything works perfect can you help me?
great, can you combine mediapipe with yolo to detect multiple objects in 1 camera, i'm looking forward to it!
Hey thankyou for the video. Is it possible to implement this code in kivy to build an android application.
Hmmmm, will dig into it.
omg love your Video thanks and like the way you code it is so clear
Hi Nicholas! First of all thanks for amazing tutorial. It was really helpful. I love your tutorials so much you are explaining the import details in very understanding way. But I want to ask you a question. I did not really fully understand the angle calculation part and I could not find anything about it on the internet either. Could you please give a little bit more info about that part?
Definitely, take a look at this: manivannan-ai.medium.com/find-the-angle-between-three-points-from-2d-using-python-348c513e2cd
Great Video !
Is it possible for mediapipe to be used in customized video (ex. not in human pose video but in other animal video)
Not this model, this one is specifically designed for use on human bodies!
my joint angle is coming near my chest and not on my elbow. the angles are a bit wrong as well. any help?
Sir , I am getting an error at angle part which you have mentioned as the angle is not showed when we on the webcam so can you help me for this ?
What's the error?
@@NicholasRenotteI have the same issue, it doesn’t show an error, but the text of the coordinates doesn’t appear in the camera
Why do you code image flag writable exactly??
Spoke to a mediapipe developer apparently it stops Python from making a copy of the frame, this improves performance.
Can you please give me a little help?? question 1: If i am just plying a video of a person and i want to visualize the coordinates next to the respective joint, what is the code i use? could you write it here? with the webcam you put cv2.puttext (image, str(angle), tuple...i guess my doubt concerns the multiplication of 640, 480.. question 2: if i want multiple joints together do i just keep listing the #get coordinates without separating anything? in other words i would want both hips, both shoulders, both elbows, knees, etc. question 3: how do i do if i want to put a true/false to turn off just the display of the angles? (im not a coder but i was able to get it all running because of your teaching, so thank you a billion!!! Good job on it!!)
Heya @W Gallo, I think I might have hit you up via email. Let me know if you're still having trouble my guy!
@@NicholasRenotte If i email you a file could you take a look at it for me?
Hello Nicholas,
Great work doing all of these; it was enlightening. I wanted to know if you have done the multi-pose/ multi gym tracking you spoke of around 52:11.
Thank you.
There's a vid on the channel somewhere, I just did the estimation not the pose detection!
@@NicholasRenotte Thank you so much! Is there a way to estimate if specific key points in a video fed into the MediaPipe are missing? Say a video of me with just my head and shoulders, and I want to calculate the key point of my ankle. Is this possible?
hello thank you for this awsome video, how we can use this code in google colab
thanks
sir can you please please tell , just like u calculated the angel using the three cordinates, how can i measure the distance between two cordinates ??
For relative distances you can use Pythagoras' theorem. For real distances I haven't dug into it.
Appreciative for your effort to teach me a comprehensive about python especially mediapipe this tool help me too much 😜😜 But nick what about hand tracking and caculate the angle between wrist and thumb, i have been tried however it not work 🥲
You got it, planning for it this week @mon mon!
@Nicholas Renotte while running make detection code, camera window pops up but after pressing "q" button to close camera window my notebook's kernel shows error that kernel is died and it gets restart and then I can't go further because of kernel died again and again. What's the issue.?
Got an error in the terminal or in your notebook?
@@NicholasRenotte sorry for not keep track of my notifications. the issue was resolved by making some version controls for some libraries, now my question is, if i want this real-time application to show me the text"XYZ" when right_shoulder, right_hip and right_knee make 90-degree angles?
I love it very much! Is it possible to calculate how accurate our poses are in order to make sure that our poses are correct comparing with gym trainer ? Thanks!!
Definitely, I'm working on a video on that rn @Tsang Wing Ho! Stay tuned!
i am getting this error even though i have install all dependencies correctly
D:\PrOjEcT\PoseEstimation>python PoseEstimation.py
Traceback (most recent call last):
File "PoseEstimation.py", line 12, in
cv2.imshow('Mediapipe Feed', frame)
cv2.error: OpenCV(4.5.1) C:\Users\appveyor\AppData\Local\Temp\1\pip-req-build-1drr4hl0\opencv\modules\highgui\src\window.cpp:651: error: (-2:Unspecified error) The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Cocoa support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script in function 'cvShowImage'
just uninstall the old opencv-python by pip uninstall opencv-python and reinstall by pip3 install opencv-python
When I tyoe the part that includes 'with mp_pose.Pose(min_detection_confidence=1, min_tracking_confidence=1) as pose:' I get this error, 'RuntimeError: ValidatedGraphConfig Initialization failed', is this project able to be completed on the CPU of a laptop and if so do you know of any solutions to my error, i have looked on google and couldnt fins a solution
i want to develop a project can i get your details
Any advice on how someone would use this method to track treadmill steps? Awesome video, BTW, very engaging!
Hi Love your work and can i see you code in github
Definitely, here it is: github.com/nicknochnack
Hey Nicholas, this is my first time using mediapipe. Why am I not getting the medipipe feed when I "print(results) in the "Make Detection section of the code (i am using Google Colab)?
Could I use MediaPipe for commercial purposes?
Double check their license terms @Khaldon github.com/google/mediapipe/blob/master/LICENSE
How does camera plays a role in Open-pose detection. Because my camera (laptop-webcam) is of low Mega-pixel, what if i use smartphone camera instead. How does it affect the pose-estimation. Thank you
Is there any way to get a somewhat workable hand tracking working on a pi zero or even a microcontroller like a esp32? Im struggling to find any real approaches and starting to think that there are none, but i dont want to give up yet.
Mediapipe on the pi is not running, even after adjusting the os version and installing hundreds of apparently missing packages
Installing the python version on an esp32 is not possible since thers only micropython and it doenst support these high-level-packages.
Since many examples for the esp are written in c++ i had hoped to be able to use the c++ version of mediapipe, but installing this is quite the opposite of straightforward and i dont even know where to start.
Every project i found with has hand tracking and a esp32 in it, makes use of the .js version where the calcutation is outsourced to an client
Thats not an option for me
Am I chasing ghosts here? Is that simply not possible? Even with very few frames?
Thanks for your great tutorials. Is there any tutorial, that shows how we can customize/modify or retrain the Mediapipe models?
Current models are trained for adult pose detection and have issues in pose detection in infants.
How can I improve it?
Hey friend...this is AMAZING!!!! Can you help me with a question??? im following your steps but cannot go beyond step 2. After telling me it successfully installed mediapipe, after the first step, it tells me that it cannot recognize this line: "import mediapipe as mp" ?? what am i doing wrong?? could you please enlighten me?? Thank you for any help!!!
Heya! Oh sometimes it doesn't pick up straight away. Try restarting your notebook and you should be good to go!
@@NicholasRenotte Hi Boss, please can you help me? I can not import cv2, i m a beginner :)
How Coordinate to visualise a specific body parts while using openCV . Example - elbow (640. 480)
Hello when trying to use mediapipe in jupyter lab I run the error "module 'mediapipe' has no attribute 'solutions' " Do you have any clue how to fix this?
Sir after installing mediapipe using pip, when I import it in jupyter notebook, it gives me an error of module found, how can I resolve this??
Thank you!!, Your video is really helpfull to me, When I make pushup counter, I subscribe you and push the good button!
YES YES YES!!! The video I more than needed. So helpful! Thank you!!!
If i wanted to use this method to calculate work done (WD= Force x Distance), how would you suggest I go about it. My thinking is that I'd assign a certain weight to either the joints or the line between joints and use that for force (signifying weight distribution in the body) then the model tracks the distance each joint or line moves throught space and multiplies to get work done. The individual work done per joint or line is summed and I'd have total work done by the body. Practicalizing this is what I can't seem to figure out.
I actually looked into this a while ago, it'd probably be something along the lines of angular force. Take an assumption of the mass of the arm, then calculate it's accleration (speed between frames) then assuming you know th length of the arm or use a proxy e.g. coordinate to coordinate euclidian distance...you should be able to do it!
You want to use that work done found as an estimate to Joules of energy used?
May I know which webcam you are using. Btw, love your content
Thanks a mil @Venkat, I'm using a Logitech StreamCam
Hey are there any references, how can I add more exercises in it? and they all should work simultaneously. Thanks!
may i know in which app to write the code? thank you
Jupyter Notebooks :)
@Nicolas Thanks it is such great help. I have 2 quick questions
>>>> why have you multiplied in angle code np.multiply(elbow, [640, 480]).astype(int)
>>>> how to resize the video frame if you can suggest some learning sources on same. It will be wonderful. Big thanks again man
Heya @Karamvir,
1. This is because we receive normalized landmarks we need to return them to our frame size to render them based on our baseline image
2. Can resize it by adding cv2.resize before the render :)!
@@NicholasRenotte can you give the landmark for all body part such as knee , shoulder to render in our frame
Hi I used openpose to get json keypoints files can you do a tutorial how to converte this json files in a 3D files as BVH or else? please
Nice tutorial!. Great!. Do you know if you can use this for multiple dector pose for people at the same time? How can i get it?
Heya @TechTales SC, this model doesn't support multiple people. Check out OpenPose for it
Hi Nicholas
I need to know what kind of data type is the holistic result and how do I save only the skeleton in a local folder.
Thanks
hey guys, did anybody else have this attributeError: module 'mediapipe.python.solutions' has no attribute 'Pose' ?
Can you tell me how to solve a parameters mazes with phyton tensor flow And ai gym
Heya @Sangita, check this out: www.samyzaf.com/ML/rl/qmaze.html
You wont read this, but. I notice that the landmarks coordinates are in 3d, but you calculated the angles just for 2D, why is that?
Like, I don't understand why in this formula you do not use the index-2 coordinates of the points:
radians = np.arctan2(c[1]-b[1], c[0]-b[0]) - np.arctan2(a[1]-b[1], a[0]-b[0])
Visuale angle and counter logic not working write the code with PyCharm, could you please help :) if you send me .py version I would be very glad to you. Btw I use Mac but I'm going to make the project to use Jetson
Weird, what's happening? You can run this inside of PyCharm just gotta save the notebook as a Python script :)
@@NicholasRenotte Hi, I solved it, totally my fault :) I was forgot the "def" :D so it could not calculate the angle :) Thank you so much, the code is very useful for our project. By the way, our project is about the physical activity at the bus stops, with this code we will guide the people during waiting. If they complete the movements, system will donate one turkish lira to ALS foundation. Thank you again, be safe and take care.
@@MyBACHCHE that is soooo cool. Do you have a video of it working? I would love to see it and share it with your permission!
May i ask what algorithm is/are involved in Mediapipe?
hii thanks your video help me finding the cordinates of joints
any way you could remake this video with the new media pipe?
Sir, can you upload tutorial for pose correction as well??
Nicholas please make us video how to detect steps
Hello Sir, If i only want specific landmark not all landmark then what change i have to do??
How can i get the code the link doesn't work?😢