IMUDINO: Self-distilled Foundation Model for Inertial Sensing
The paper presents IMUDINO, a foundation model for inertial sensing that achieves competitive performance across diverse sensing tasks.
The paper presents a novel application of self-supervised learning to inertial data and a new model architecture for inertial sensing.
Before reading this…
Applications
- →Human activity recognition
- →Health monitoring
- →Motion capture
- →Odometry
To understand this paper, make sure you know these concepts first:
- Machine learningfind papers →
- Self-supervised learningfind papers →
- Transformer encoderfind papers →
Abstract
More Like ThisWe present IMUDINO, the first foundation model (FM) for inertial sensing to achieve competitive performance across diverse sensing tasks including human activity recognition (HAR), health, motion capture, and odometry. Prior inertial FMs often focus on one subfield of inertial sensing and sometimes underperform state-of-the-art supervised methods (e.g., in HAR), making their scope considerably narrower than FMs for vision and natural language. Moreover, they often require fixed input length and sampling rate, and generalize poorly to unseen sensor locations. To build a general-purpose foundation model for inertial sensing, we adapt the DINO self-supervised learning (SSL) paradigm to inertial data through a set of semantics-preserving motion transformations that simulate variations in motion pattern and sensor setup. For model architecture, we use a plain transformer encoder with frequency-aware positional encoding that natively supports varying input length and sampling rate. This simple setup proved remarkably effective for learning global and dense features generalizable across users, sensor setups, tasks, and, in the case of robot odometry, embodiments. We hope IMUDINO can serve as a backbone for inertial sensing and kickstart the development of general-purpose inertial FMs.