Within the framework of the Social Wear research project, we have developed an innovative approach to motion capture (MoCap), where capacitive sensors are integrated into fashionable, loose-fitting clothing. These sensors, made from conductive textiles, enable continuous capture of dynamic upper body poses.In contrast to conventional Inertial Measurement Unit (IMU)-based MoCap systems, which rely on inverse dynamics, MoCaPose decouples sensor positioning from pose capture. By utilizing a deep regressor, precise 3D coordinates of body joints are predicted regardless of specific applications.
Our project underwent several prototyping iterations to address technical challenges and perfect textile integration through fashion-technological co-design. Extensive datasets, comprising synchronized video and capacitive data from various participants, were used for validation.
Furthermore, as part of the project, a jacket was developed that serves as a communication aid for cyclists. Based on the requirements of the research consortium of the involved DFKI departments and the research questions to be examined, we worked on the implementation using the MYOW toolkit 3.0. The result is a jacket equipped with three sets of three vibration motors integrated in a matrix-like arrangement in the upper body area. Additionally, the jacket features an LED matrix on the back that communicates signals and information to other road users.
The jacket is intended to be controlled via an app specifically developed for use in cycling swarms. The goal was to investigate how the swarm behavior of cyclists can be improved through minimal communication to achieve harmonious group behavior.
Involved research groups at DFKI: DRX, COS Berlin, ACG.
Researcher: Esther Zahn, Emil Woop, Clara Gleiß / DFKI DRX
Social Wear
Within the framework of the Social Wear research project, we have developed an innovative approach to motion capture (MoCap), where capacitive sensors are integrated into fashionable, loose-fitting clothing. These sensors, made from conductive textiles, enable continuous capture of dynamic upper body poses.In contrast to conventional Inertial Measurement Unit (IMU)-based MoCap systems, which rely on inverse dynamics, MoCaPose decouples sensor positioning from pose capture. By utilizing a deep regressor, precise 3D coordinates of body joints are predicted regardless of specific applications.
Our project underwent several prototyping iterations to address technical challenges and perfect textile integration through fashion-technological co-design. Extensive datasets, comprising synchronized video and capacitive data from various participants, were used for validation.
Furthermore, as part of the project, a jacket was developed that serves as a communication aid for cyclists. Based on the requirements of the research consortium of the involved DFKI departments and the research questions to be examined, we worked on the implementation using the MYOW toolkit 3.0. The result is a jacket equipped with three sets of three vibration motors integrated in a matrix-like arrangement in the upper body area. Additionally, the jacket features an LED matrix on the back that communicates signals and information to other road users.
The jacket is intended to be controlled via an app specifically developed for use in cycling swarms. The goal was to investigate how the swarm behavior of cyclists can be improved through minimal communication to achieve harmonious group behavior.
Involved research groups at DFKI: DRX, COS Berlin, ACG.
Researcher: Esther Zahn, Emil Woop, Clara Gleiß / DFKI DRX
Foto: Esther Zahn