Improper alignment during exercise can result in significant injuries, especially when people participate in home workouts or online fitness programs without professional guidance. To tackle this problem, we suggest a real-time system for correcting workout posture that leverages advanced pose estimation technology via MediaPipe to evaluate human body positioning. By measuring joint angles for a range of exercises—including squats, push-ups, planks, and yoga poses—the system assesses whether the user's form complies with ideal biomechanical practices.

The platform provides immediate auditory cues to assist users in modifying their posture during workouts, thereby reducing the risk of injury and encouraging healthy exercise routines. An easy-to-use web-based dashboard, built using FastAPI, OpenCV, and SQLite, incorporates important functionalities such as user registration, session tracking, and performance analysis. This design guarantees that data is effectively stored and accessed, allowing users to track their development over time and stay motivated.

Additionally, the system's quick responsiveness and ability to provide real-time corrections make it a viable, cost-effective, and easily accessible substitute for conventional in-person fitness coaching. Its combination of computer vision, streamlined backend services, and database support presents a comprehensive and intelligent solution for at-home fitness training, improving safety, engagement, and long-term health outcomes.