Description
Kinetic Echoes centers on the use of machine vision, specifically through the application of pose recognition technology. This system identifies human figures in images and videos by deconstructing the body into a series of key points, commonly referred to as landmarks. Utilizing advanced machine learning algorithms, pose recognition estimates the position and orientation of these landmarks in real-time. As the dancer performs, the system dynamically detects and tracks 33 distinct landmarks distributed across the body, capturing the intricate nuances of motion with precision and immediacy.
The process of pose recognition serves as the foundation for data collection in this work. Each movement of the performer is transformed into a data point—a fragment of a larger, dynamic portrait of human expression. When aggregated, these data points form an evolving depiction of motion and emotion, rendered within the digital landscape of a point cloud. This application of machine vision bridges the physicality of human movement with its digital representation, connecting the tangible and the virtual enviroments.
Mediums
Title: Kinnetic Echoes
Year of Creation: 2022
Technologies: Touchdesigner, Google’s MediaPipe Libraries, Python
Type: Generative Artwork
Print Materials: FA Archival Paper Photo Rag Metal, Wooden Frame, Glass
