I tried to build a Machine Learning Python App in 15 Minutes | Coding Challenge
HTML-код
- Опубликовано: 6 сен 2022
- What's happening guys, welcome to the fourth episode of CodeThat!
In this ep I try to build a machine learning app to track deadlifts all done using nothing but Python. In this case we used a Scikit-Learn model, mediapipe and Tkinter to get it all done.
Get the Code: github.com/nicknochnack/mlapp
Oh, and don't forget to connect with me!
LinkedIn: bit.ly/324Epgo
Facebook: bit.ly/3mB1sZD
GitHub: bit.ly/3mDJllD
Patreon: bit.ly/2OCn3UW
Join the Discussion on Discord: bit.ly/3dQiZsV
Happy coding!
Nick
P.s. Let me know how you go and drop a comment if you need a hand!
#machinelearning #codingchallenge #tkintertutorial #python 
Наука
And as a software developer myself... I gotta say I love your attitude: how can you keep such a good mood in the middle of a nearly-impossible deadline? Truly admirable
Reading this made me smile, idk man, I have no idea how I stay sane during these 😂 but I’m super happy you’re enjoying them ✌🏽
Love it. I've just discovered this channel by accident, and I'm watching one video after another. I suddenly feel like coding along.
With your videos my learning curve will shrink drastically. Thank you.
Fantastic!! Nicholas I Am a big fan of you. I love Machine Learning and your contents helps me to learn more and more in this beautiful area. Thank you so much for your videos!!
No doco, no stacko. Most Australian dev ever 😂
😂😂😂😂 "And for the penalty it's $50 casho!"
Thanks Nicholas! Following your video lessons makes me learn really really much on the detection machine learning skills~ May I suggest if you can come out with one teaching video on data augmentation on the Mediapipe keypoints (e.g. shift, rotate, zoom in/out etc.) so that we can have more comprehensive dataset to train~ Thank you! Hope u can see my comment from all ur notifications lol!
Fantastic! It must be hard to code so fast and explain at the same time. And this trick with multiline editing looks really cool, i didnt think its possible to copy-paste using this
How good is it right?! My new favourite VS Code hack
@@NicholasRenotte Man something like this was needed so much, and I didn't find anywhere else, suddenly your video popped up! Thanks, it's realllyyy goooddd
@@NicholasRenotte I tried to google it..but everytime the cursor doesn't come at the end of the line....what keys to use to get all the cursors at the end of the line
It is possible to copy/cut/paste multiple lines and have them land where you want. I know this works in Webstorm, idk about Visual Studio, but If you use 3 cursors to cut 3 lines, then elsewhere, put 3 cursors and paste, you should get the results you expect :)
Doesn't that kinda defeat the DRY principle though?
I love itttt, you are really great at coding. Let me introduce myself, I am doing bachelors in computer science, currently in last year , and I am doing part time job as a data scientist, but I like to explore different perspectives.
Awesome stuff @Hardy!! Welcome!
This was so fun to watch omg hahah , keep dem videos coming !!!
😂😂 cheers @Chris!
you and your videos(knowledge) are amazing, I wish one day I can be like you in programming
Thanks @Mehran!!
Great! It's an exciting challenge that keeps the audience engaged and entertained. Keep up the fantastic work! :)
awesome stuff. ended up adding a beep feature after each successful move up!!
You are amazing though. Learned a lot and got lots of encouragement from you. Thanks so much for your videos.
Thank you so much for checking them out @林世奇!! Glad you're enjoying them.
Amazing! Thank you for the great idea and video.
Thanks a mil @Simon, stoked you liked it 🙏
This is awesome stuff!! Really looking forward to next video
For the GUI I use QT designer. You don't need to Google the commands. The only problem is you need to design the whole GUI than code the backend.
Whhhaaat, that's actually sick. I was randomly trying to work out how to get it work when I first started. Tbh I want to start doing some more stuff in Js but thanks for the heads up!!
This was so amazing to watch! Thanks!
Cheers @AI Manifest!
Hi Nick...this was a brilliant video. Any chance you could make a long form video tutorial around this? I loved the RL tutorials with SB3
Will try to get back to it!!
omg awesome application bro keep doing apps like this
Excellent brother...love from India ❤
Absolutely amazing video! By any chance, do you have a video that goes through how you built the deadlifting model?
Check this out @Christopher: ruclips.net/video/PGsAsuwBdw0/видео.html I can't remember if I shared the code for the model building or not, if I didn't let me know!!
@@NicholasRenotte I don't see it in your github or on the link for the video.
So few time and so many informations. Thanks guy.
Anytime 🙏 tried to pack it in!!
Mannnn !!!!!... This is so adrenaline.... just like watching penalty shoot-out of a football match ....Thanks for making the process of learning engaging way ....
GOOD JOB !!!!! I KNOW I WOULD NEVER be able to even do something like that...
Really sickstuff man!
Cheers HyperDoge!
Project idea:
Is it possible for the camera to locate the position of the eyes and move a mouse cursor? The right eye blink would be the right button and the left eye would be the other. It would be a great application for paraplegic people.
Ooooooh, I like this. Let me see what I can do!
@@NicholasRenotte So did you mean " Yes. Possible" ? :)
extremely marvelous idea ,hope that it will be possible to build.........
@@mehmetsarsin This is available online in websites
Funny enough it's what iv been working on!
Damn, I was thinking about that yesterday app that could count calories and reps, series. Log data and monitor progress. This is amazing.
Marvellous job boss ❤.
Hi Nick, great work truly. I love watching your videos. It has helped me gain a huge heap of knowledge into building models for my assignments. I really look up to you as a mentor and you are just awesome.
How about delving into autonomous navigation and reinforcement learning?
Probs need more than 15 minutes for that but I plan to get back into it when I get a new PC, think i need a beefier GPU to handle it!
@@NicholasRenotte Yes kinda hoping for a whole course on that. Cause you are so thorough with your videos it is so easy to gather concepts to go ahead and start building. Looking forward to your upcoming videos 😊
Could you make another 5 hrs video explaining this 15 minutes? 😉
You're amazing bro
🤣
LOL maybe someday man 😂 thanks for checking it out 🙏
Bro, i love your character, keep goin
🙏 🙏 🙏 thanks @afshin!!
Thanks for the help.
Hi Nick, brilliant ideas and love your videos. Is there a resource to how the model itself ? I'm kind of a newb to ML to I'd like to learn that as well :)
Nice video , i have a question for you : do you have strong maths background i mean did you understand at all the book deep learning of ian goddfellow and yoshua bengio ?
awesome videos as usual
Thanks a mil @Kalid!!
Hey here is a project idea for you/anyone that wants to do it.
Use the AI to check the form of different work outs thatway people can check to make certain they are doing it right if they happen to be lifting alone/for the first time or if they just want to get stronger.
I always wanted but don't have coding skill
theres heaps of issues with this, one is bracing and the other muscle contraction. Can a video simply see those things? what if your core is covered by a tshirt, how can the camera see all angles. Need more than just one angle, when deadlifting, is the chest up, is the core braced etc but yes I pondered on this thought too
Thanks a ton!
it was really a good project, i tried on my first laptop, it show me a lot of errors, again i change to another laptop and it works, i finish it maybe less than 40minutes with the video,
thank you
WOAH nice! Awesome work!! Most of the time it's mediapipe that can be a bit finicky
Loved it ❤
dude this is bad ass should i try it later?
Hi nick, i am just starting my machine learning journey. I have my college project this year and i decided to work on sign language detection . I'm having a really hard time installing things in the anaconda prompt. What are all the packages i will have to install. And in what order should i watch your videos of sign language detection. Desperate for help.
Thank you for the information and the video. I have few issues running the code. The most important one is : what is 'deadlift' ? is this a dataset on your device or part of the pickle module ?
yeah me either , i stuck here ..
Thats impressive !!
couldn't find the landmarks library instead i found landmark-detection so it is import as import landmark as LM . will that work?
encounntered same problem
bro where chalk up u been, it is so cool
LOL, been here man!! Thanks a mil!!
this is truely amazing !
can you please share the dataset that you used in landmarks, I'm having difficulty in it. Thanks for making this video.
How did you imported all these libraries as easily at the beginning, I'm trying to learn computer vision however I can't follow 10 minute of my course before VS Code tells me that there is a problem. Could you please make a simple video about how to create a folder with all the stuff to follow a computer vision course with mediapipe and cv2 and how to make it works (i always have problems in the console and if not, when I try to run my webcams, my webcams appears to be black). Could you please make a short but complete video about it. 🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏
Great Nick, still waiting for an app where u can use an iPhone as a camera😉
Learning React Native as we speak!!
Vislibity is indeed an hell of a word
if anyone of you are building with customtkinter 0.3 version, then you can directly use text parameter in CTkLabe like this CTkLabell(text=""), don't need .configure(text="").
is it the reason why i am getting tcl text unknown option
and how to solve this with last three lines of code where he used .configure
I wish you would also include making the menu and pages....but thats asking too much :)
Thank you!!
Why is the image captured by the camera at an angle when I run this program? How can I adjust it?
What are pre requisites, I have done python and library (numpy , pandas ) can I directly start this tutorial
Great Work! I appreciate the clean and clear videos you make explaining the process. Could you do a video about SLAM or PyTorch3D? I've been having difficulty finding good videos that implement the code that I can follow along. Thanks!
Ooooh I've wanted to get into point cloud stuff, let me take a look!!
Dose your projects need strong features of laptops i have i5 20 RAM
could you please make a video about the startup environment of this code I have your code and also downloaded all the libraries but still can not run it.
Hi nick, can you share the jupyter notebook of code on how you actually built the deadlift.pkl model . Ty
I join to the request - would be nice to see how you built this module. Please share it. THX
Even if you give away a coupon I won't redeem it so you don't end up spending too much. I'm already indebted to you for all the things you did for learners like me. Thank you, Nicholas.
🙏 🙏 🙏 Thanks Taeef!!
@@NicholasRenotte You're welcome! and thank you very very much again
Can you make this type of video for implementation in Android app... I am really excited to see that implement ml model with a simple Android app
Me too!
An-mhaith (Well done) Pal, nice one.
Thanks a mil @Muhammad!!
Why does the GMS tNice tutorialng doesn't soft sa way like it did in the video?
If we are create sign language app using this model source code help what should changing in my project? Please quickly answer its very urgent.
hi , can some how i use this code for builting andriod app is this possible . plz reply .. i am excited
what python version are you using, please ? (:
error issue: i have download this code from your GitHub but i can't run it, the error said that i have problem with my "tkinter" and "customtkinter" library even though I have install it with pip in vscode
Can you show something similar with your previous sign language model
Big up big Nic! That's how you supposed to code .
Thanks a mil Swapnil 🙏
@@NicholasRenotte that rhymed 😊😅
what if i use hdf5 file to save my model? what should i use?
this is your only project that close to commencial use.
interesting
Hi Nick. App is very cool. Are you try app like this to do in unity engine?
Honestly I'm up for anything these days. I'm looking at Flutter and Rust projects next, but will add on Unity. Cheers for the suggestion Dragan!
could you make a video on identifying exterior car parts and then identify the damages
Sure!! Got any datasets for me I can take a look at?
@@NicholasRenotte i have been trying to post the links but im not sure if youre getting any
Yo nic, can we have a full project for TF ViT. Currently there is a lack of ViT tutorials everywhere.
You got it! I've got a full tutorial coded up but never released it!
@@NicholasRenotte cant wait 😍😍
Hey Nick... It says "ModuleNotFoundError: No module named 'landmarks' .... I also tried doin "pip install landmark-detection". It doesn't work for me . help pls
Should be .landmark :) like so: results.pose_landmarks.landmark
please make video on
how to integrate driver drowsiness ml project to andriod and ios app?
Hey Nicholas, iam an engineering student
iam trying to create sign language video calling application for my major project. can you help me out with it? For some knowledge how to stream the video
Or to link it to a calling application
It will be really grateful if you just give me some tips
Insane stuff... 🙂
I would like you to teach me how to create pickle files for different exercise postures.
Heya @Parinya, check out the live stream I did on it!!
what webcam do you have?
Exception has occurred: AttributeError
partially initialized module 'cv2' has no attribute 'gapi_wip_gst_GStreamerPipeline' (most likely due to a circular import)
How can I solve this
Haven't seen that one but looks like an opencv error, possibly try reinstalling?
Dear Nicholas the video was great but I couldn't figure it out by importing landmarks , I have tried it thousend times but I got still nothing. The Error is: ModuleNotFoundError: No module named 'landmarks'. So I need your !!!!HELP!!!!
I think you could have won if you didnt explain, but thank you very much as always!
Thanks a mil @bonvivant, can't sacrifice explaining it though!!!
Thank you for the interesting video and the Amazon gift card:)))
I saw your message via Amazon!! So glad you got it! Keep at it Alexey!
How to create that deaflift.yml ? please
Yoh!... please how did you do the multiple cursors😮
Hold Alt and click where u want the cursor
Sick right?! Also if you select a variable or value hit Alt + D multiple times to select multiple.
@@NicholasRenotte yay!! Worked..I'm so pumped up right now.. just like @palgun said, holding alt while clicking on any point does the magic.. pretty awesome.. thanks bro
YESSSSS! Lessssgoooo!!
I tried your code but it shows [Errno 2] No such file or directory
Is this python 3.11 or older? Since Mediia pipe not working for me in 3.11.
Same error man
can u please share other exercises pkl files....
React native can this be done with that ?
hello sir.... can you make makeup transfer model using neural style transfer
Can this be done with react native
Official surpassing Andrew ng in my heart in terms of respect
Woah woah woah....Harry....what?!? 😂 Thanks so much 🙏
@@NicholasRenotte yea I’m learning so much from ur vids in terms of implementation and inspiration for some of my own project ideas that Andrew Ng’s lectures don’t provide
'str' object has no attribute 'predict_proba' in pkl file
It's ironic that the video related to computer "vision" has the problem with poor visibility at the word "visibility".
LOL, yeah it kills be that I even called it out 🤦♂️
@@NicholasRenotte you're doing great btw. Keep up the good work.
So cool
I'm getting this error _tkinter.TclError: unknown option "-text"
I'm new here so how can we use multiple models in this???
You can bring more in where I added the pickle load. The implement the same (or similar logic) for form tracking. I showed it in action here: ruclips.net/video/H7cGq0xIHbc/видео.html
@@NicholasRenotte Thank you for helping
Can we have EDA series?
You got it!
@@NicholasRenotte thanks 👍👍
the pickle file doesnt work??