You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
MoveNet-Web is a web application that leverages TensorFlow.js and the MoveNet model to perform real-time pose estimation directly in the browser. Built with Next.js and utilizing the WebGL backend, this application ensures efficient GPU acceleration for optimal performance.
This project develops a fall detection system using Pose Estimation and Optical Flow data with LSTM networks. It enhances detection accuracy for elderly care by analyzing body movements in real-time. Key technologies include the MoveNet model for pose estimation, Optical Flow analysis and LSTM networks for temporal analysis.
This system uses YOLO for object detection (specifically, garbage), MoveNet for hand landmark detection, and DeepFace for facial recognition. It analyzes the relationship between detected humans and garbage to identify potential littering incidents in real-time.
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.