Magnitude and Rotation Invariant Detection of Transportation Modes with Missing Data Modalities
This work provides an incremental improvement for transportation mode recognition systems that must handle missing sensor data and distribution shifts.
The authors tackled transportation mode detection using phone sensor data with randomly missing modalities, addressing distribution shifts between train and validation data. Their rotation-invariant aggregation method achieved substantial performance improvements over baseline approaches while reducing feature vector length.
This work presents the solution of the Signal Sleuths team for the 2024 SHL recognition challenge. The challenge involves detecting transportation modes using shuffled, non-overlapping 5-second windows of phone movement data, with exactly one of the three available modalities (accelerometer, gyroscope, magnetometer) randomly missing. Data analysis indicated a significant distribution shift between train and validation data, necessitating a magnitude and rotation-invariant approach. We utilize traditional machine learning, focusing on robust processing, feature extraction, and rotation-invariant aggregation. An ablation study showed that relying solely on the frequently used signal magnitude vector results in the poorest performance. Conversely, our proposed rotation-invariant aggregation demonstrated substantial improvement over using rotation-aware features, while also reducing the feature vector length. Moreover, z-normalization proved crucial for creating robust spectral features.