Skip to content

Latest commit

 

History

History
57 lines (36 loc) · 1.51 KB

README.md

File metadata and controls

57 lines (36 loc) · 1.51 KB

MoveMentor 💃🏻

When Movement Meets Motication!

Setup 🛠️

Pre-Requisites: Flutter >= 3.7

  1. Install TFLite dependencies for Android

Linux / MacOS

sh install.sh -d

Windows

install.bat -d
  1. Install TFLite dependencies for iOS
  • Download this file: TensorFlowLiteC.framework
  • Place the TensorFlowLiteC.framework file in the pub-cache folder of the tflite_flutter package
  • Linux/ Mac => ~/.pub-cache/hosted/pub.dev/tflite_flutter-/ios/
  • Windows => %LOCALAPPDATA%\Pub\Cache\hosted\pub.dev\tflite_flutter-\ios\

** To use Metal on iOS: Download the file from this PR: https://github.com/am15h/tflite_flutter_plugin/pull/200/files and follow the instructions here: https://github.com/am15h/tflite_flutter_plugin/wiki/Build-iOS-binaries-with-GPU-delegate

  1. Install Flutter Dependencies
flutter pub get
  1. Run
flutter run

Model Specifications 👨🏻‍💻

  • Using the MoveNet Lightning Single Pose Model
  • Uses a 640x480 RGB image as input (Medium Resolution Preset for Android and iOS)
  • Outputs an array with [x, y, confidence] (x, y are scaled according to the input image size)

Similarity Calulcation

Angle Between Two Lines: https://www.nagwa.com/en/explainers/407162748438/