Summary of Generalizable Sensor-based Activity Recognition Via Categorical Concept Invariant Learning, by Di Xiong et al.
Generalizable Sensor-Based Activity Recognition via Categorical Concept Invariant Learning
by Di Xiong, Shuoyuan Wang, Lei Zhang, Wenbo Huang, Chaolei Han
First submitted to arxiv on: 18 Dec 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- 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 This paper addresses a crucial issue in Human Activity Recognition (HAR), where models struggle to generalize across different test sets due to inter-subject variability. The authors propose Categorical Concept Invariant Learning (CCIL), a framework that regularizes the model during training by introducing a concept matrix. This matrix encourages feature-invariance and logit-invariance, enabling the model to recognize activities in unseen distributions. CCIL outperforms state-of-the-art methods on four public HAR benchmarks, demonstrating its effectiveness in cross-person, cross-dataset, cross-position, and one-person-to-another settings. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research paper is about making machines better at recognizing human actions, like walking or running. Right now, these machines can be really bad at understanding what people are doing if the person is different from the ones they were trained on. The problem is that humans come in all shapes and sizes, so it’s hard to make a machine understand what everyone does. The paper introduces a new way to teach machines called Categorical Concept Invariant Learning (CCIL). It helps machines learn to recognize actions in situations where people are different from the ones they were trained on. |
Keywords
» Artificial intelligence » Activity recognition
