Summary of Flow: Fusing and Shuffling Global and Local Views For Cross-user Human Activity Recognition with Imus, by Qi Qiu et al.
FLOW: Fusing and Shuffling Global and Local Views for Cross-User Human Activity Recognition with IMUs
by Qi Qiu, Tao Zhu, Furong Duan, Kevin I-Kai Wang, Liming Chen, Mingxing Nie, Mingxing Nie
First submitted to arxiv on: 3 Jun 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 The paper proposes a novel approach for Human Activity Recognition (HAR) using Inertial Measurement Unit (IMU) sensors, addressing the challenge of achieving robust generalization performance across diverse users. The main issue is the variation in data distribution among individual users due to the representation of IMU sensor data in the local coordinate system. To alleviate this issue, a global view representation is extracted based on the characteristics of IMU data, which is then compared with local view data. Experimental results show that the global view data outperforms local view data in cross-user experiments. Furthermore, a Multi-view Supervised Network (MVSNet) is proposed to fuse local and global view data, using shuffling and view division to supervise feature extraction. The paper demonstrates state-of-the-art performance on OPPORTUNITY and PAMAP2 datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The researchers are trying to make computers better at recognizing human activities, like walking or running, by using special sensors called IMUs. These sensors are really good for this job because they’re small, don’t use much energy, and lots of people want to use them. The problem is that these sensors work differently depending on how people wear them, which makes it hard for computers to recognize activities accurately across different people. To solve this problem, the researchers came up with a new way to look at the sensor data that ignores how people wear the sensors. This new way works really well and beats all other methods at recognizing activities. |
Keywords
» Artificial intelligence » Activity recognition » Feature extraction » Generalization » Supervised
