Summary of Data Augmentation Techniques For Cross-domain Wifi Csi-based Human Activity Recognition, by Julian Strohmayer and Martin Kampel
Data Augmentation Techniques for Cross-Domain WiFi CSI-based Human Activity Recognition
by Julian Strohmayer, Martin Kampel
First submitted to arxiv on: 1 Jan 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
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 paper investigates ways to improve the recognition of human activities using WiFi Channel State Information (CSI) in indoor environments, focusing on model generalization performance. By applying image-based learning data augmentation techniques to WiFi CSI, the authors aim to enhance model robustness across different environmental conditions and sensing hardware configurations. The study utilizes a dataset of CSI amplitude spectrograms, training EfficientNetV2-based activity recognition models to evaluate the effects of various augmentations on model generalization. The results demonstrate that specific combinations of simple data augmentation techniques can significantly improve cross-scenario and cross-system generalization. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper looks at how to make it better for computers to recognize people’s actions in homes using WiFi signals. Right now, these systems are not very good at understanding activities when the environment or equipment changes. To fix this, researchers tried applying tricks they normally use with images to WiFi signals. They tested different combinations of these tricks on a big dataset of WiFi signal patterns from various scenarios and antenna types. The results show that using the right combination of these tricks can make the system much better at understanding activities in different situations. |
Keywords
* Artificial intelligence * Activity recognition * Data augmentation * Generalization
