Using a Stereo-Vision Setup to reliably track a Player's Elbow

This project uses the MoveNet model for pose estimation. The model is loaded and run using TensorFlow Lite. The predicted keypoints are then visualized on the input image.

The repository is to predict human joint location from JPG images that have a pixel size of 256*256. This will be done using the Movnet Singlepose thunder pretrained model, which will be deployed on AWS using CF and SDK.

Computer Vision model for identifying police hand signals using TensorFlow & Tensorflow-lite.

AI based yoga trainer.

This project recommends seven tailored asanas based on a user's health conditions. Additionally, our yoga pose estimator detects incorrect postures for improved form.

This is a project to implement a human fall posture recognition model through LSTM deep learning model using Flask web framework.

Live Yoga poses detection

This is a project to recognize human fall posture through LSTM deep learning model.

This project is an implementation of MoveNet which is developed by Google. Inspired by monolesan's fix_posture project,we are going to set more thresholds and build your personal deep-learning model less than 5 min.

A Flutter app that utlises TensorFlow's MoveNet model for pose detection, in order to count a user's reps and detect the correctness of their pose, with additional modules for BMI calculation, nutrition.

Monitor Your Workout through a Webcam/IP Camera. No equipment is required, other than a camera and a laptop. This application could potentially replace a personal trainer, making it the idea app for workout.

This project focuses on Human Pose Estimation using the MoveNet model with TensorFlow Lite. The goal is to detect keypoint positions on a person's body in images and live video frames. The project provides a Flask web application for both image and live video input, showcasing the real-time capabilities of the model.

The Tennis Serve Analysis App is a mobile application designed to revolutionize the way tennis players analyze and improve their serves. Leveraging machine learning algorithms and computer vision techniques, the app provides users with personalized feedback of their serves.

MoveNetを用いたPythonでの姿勢推定のデモ