Summary of Magnitude and Rotation Invariant Detection Of Transportation Modes with Missing Data Modalities, by Jeroen Van Der Donckt et al.
Magnitude and Rotation Invariant Detection of Transportation Modes with Missing Data Modalities
by Jeroen Van Der Donckt, Jonas Van Der Donckt, Sofie Van Hoecke
First submitted to arxiv on: 8 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The Signal Sleuths team presents their solution to the 2024 SHL recognition challenge, which involves detecting transportation modes using phone movement data with missing modalities (accelerometer, gyroscope, magnetometer). To address the significant distribution shift between train and validation data, they employ a magnitude and rotation-invariant approach. Their method combines traditional machine learning techniques, focusing on robust processing, feature extraction, and rotation-invariant aggregation. The ablation study reveals that relying solely on signal magnitude vector performs poorly, while their proposed rotation-invariant aggregation improves performance while reducing the feature vector length. Additionally, z-normalization proves crucial for creating robust spectral features. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The Signal Sleuths team solved a tricky problem to recognize transportation modes from phone movement data with missing sensors. They had to find a way to make the system work well even when some of the data was missing. To do this, they used special techniques to process and combine the remaining data in a way that didn’t rely on knowing which sensor was missing. This approach worked better than trying to use only the most common type of data or using features that took into account which sensor was missing. The team also found that normalizing the data helped create more reliable results. |
Keywords
* Artificial intelligence * Feature extraction * Machine learning
