DIY AI Golf Swing Analyzer: Leveraging Body Landmark Detection from Your Smartphone
The author developed an AI tool that tracks 33 body landmarks using standard smartphone video to analyze golf swings. They referenced existing implementations of this technology in golf swing analysis, notably by developers cited in Banze, 2025. This project demonstrates practical application of pose estimation models outside elite labs.
This story illustrates how pose estimation AI models, previously confined to research or specialized sports labs, can now be applied by individuals using everyday devices. It encourages readers to rethink their workflow by integrating accessible computer vision tools to gather precise biomechanical data without expensive equipment.
Researchers and developers at Banze Labs (2025) pioneered applying body landmark detection specifically for golf swing analysis, enabling accurate swing feedback with minimal hardware. Their work sets the precedent for DIY efforts like this author's personal project.
Step 1: Record a video of a golf swing using your smartphone. Step 2: Use a pose estimation library such as Google's MediaPipe (https://mediapipe.dev) to detect and extract 33 body landmarks from the video frames. Step 3: Analyze the landmark data to assess swing mechanics and identify areas for improvement. Expect detailed joint position data that can inform coaching or self-assessment.