mlkit
mlkit copied to clipboard
iOS CSV pose classification
I can already use MediaPipe Colab. Currently, Android has a csv posture classification and counting method. For iOS, I don’t know how to classify posture actions based on csv files. Will this part of the code be added in the future?
Thanks for your feedback! We don't have current plan to provide an iOS example, but if we did, iOS would have the same logic as Android. If you are familiar with Java, you can try to map the Android classification code to iOS code. The classification code doesn't use any platform specific things, just general process and algorithm.
Thank you. I tried Android Studio and opened the code in the demo. I think I don't know Java and it's still hard to map the Swift code
@sujie0308 is ur problem solved? can you share how to solve this issue?
I'm also looking for iOS examples
I was also looking for iOS samples, but didn't find a solution either. So I mapped the java code to swift and rebuilt the minimal pose classification example. You can find the code here: https://github.com/Throvn/MLKitPoseClassificationExample
Note though, that I am not a Swift native coder, so parts of the code could be a bit messy. Therefore any kind of pull requests (like code cleaning, or adding a readme) are more then welcome.
We don't plan to provide an iOS example, but you can use the Android classification code as an example and try to replicate it on iOS. The classification code doesn't use any platform specific things, just general process and algorithm.