Summary of A Masked Semi-supervised Learning Approach For Otago Micro Labels Recognition, by Meng Shang et al.
A Masked Semi-Supervised Learning Approach for Otago Micro Labels Recognition
by Meng Shang, Lenore Dedeyne, Jolan Dupont, Laura Vercauteren, Nadjia Amini, Laurence Lapauw, Evelien Gielen, Sabine Verschueren, Carolina Varon, Walter De Raedt, Bart Vanrumste
First submitted to arxiv on: 21 May 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 A novel semi-supervised machine learning approach is proposed to recognize micro activities in the Otago Exercise Program (OEP), a vital rehabilitation initiative for older adults aimed at enhancing strength, balance, and preventing falls. The existing HAR systems neglect the recognition of individual repetitions of exercises, focusing only on macro activity sequences. The proposed model uses a Transformer encoder for feature extraction and classification by a Temporal Convolutional Network (TCN). Additionally, masked unsupervised learning is employed to reconstruct input signals, which enhances the performance of the supervised learning task, achieving f1-scores above the clinically applicable threshold of 0.8. This approach enables automatic monitoring of exercise intensity and difficulty in daily lives of older adults. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The Otago Exercise Program helps older adults improve their strength and balance to prevent falls. Right now, machines can recognize what people are doing for a long time, but they can’t tell exactly how many times someone does an exercise or how fast they move. This paper introduces a new way to use machine learning to better understand the exercises and how well people do them. The approach is semi-supervised, which means it uses both labeled and unlabeled data. It also combines two techniques: one for feature extraction and another for classification. This helps the model learn more from the available data. The results show that this approach works well, with scores above a certain threshold that is important in clinical settings. By counting repetitions of exercises and calculating speed, this approach can help monitor exercise intensity and difficulty in daily life. |
Keywords
» Artificial intelligence » Classification » Convolutional network » Encoder » Feature extraction » Machine learning » Semi supervised » Supervised » Transformer » Unsupervised
