Summary of Sudokusens: Enhancing Deep Learning Robustness For Iot Sensing Applications Using a Generative Approach, by Tianshi Wang et al.
SudokuSens: Enhancing Deep Learning Robustness for IoT Sensing Applications using a Generative Approach
by Tianshi Wang, Jinyang Li, Ruijie Wang, Denizhan Kara, Shengzhong Liu, Davis Wertheimer, Antoni Viros-i-Martin, Raghu Ganti, Mudhakar Srivatsa, Tarek Abdelzaher
First submitted to arxiv on: 3 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 introduces SudokuSens, a generative framework for creating synthetic training data that mimics experimental configurations not encountered during sensor data collection in IoT applications. This improves the robustness of deep learning models, which are crucial for IoT applications where data collection is expensive. The authors motivate this work by noting that IoT time-series data entangles object signatures with environmental properties and disturbances. To address this, they employ a Conditional Variational Autoencoder (CVAE) to reduce data collection needs from multiplicative to linear while generating synthetic data for missing conditions. They also use a session-aware temporal contrastive learning approach to improve robustness against dynamic disturbances. By integrating these approaches, SudokuSens significantly improves the robustness of deep learning models for IoT applications. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a tool called SudokuSens that helps make machine-learning-based Internet-of-Things (IoT) applications work better. It makes fake training data that looks like real data collected from sensors, which is important because collecting real data can be expensive and time-consuming. The authors made this tool to help with IoT applications where you have lots of devices sending data in real-time. They used special techniques to make the fake data look realistic and also to make sure the models are good at handling unexpected things that might happen. |
Keywords
* Artificial intelligence * Deep learning * Machine learning * Synthetic data * Time series * Variational autoencoder
