Summary of Em-net: Gaze Estimation with Expectation Maximization Algorithm, by Zhang Cheng et al.
EM-Net: Gaze Estimation with Expectation Maximization Algorithm
by Zhang Cheng, Yanxia Wang, Guoyu Xia
First submitted to arxiv on: 11 Dec 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: 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 proposed EM-Net model uses a combination of deep learning and traditional machine learning algorithms to improve gaze estimation accuracy while reducing computational demands. The model incorporates a Global Attention Mechanism (GAM) to extract features related to gaze estimation, enhancing its ability to capture global dependencies. Additionally, the Expectation Maximization (EM) module enables the model to learn hierarchical feature representations, promoting generalization and minimizing sample size requirements. Experimental results demonstrate that EM-Net outperforms GazeNAS-ETH on three datasets (Gaze360, MPIIFaceGaze, and RT-Gene), achieving improvements of 2.2%, 2.02%, and 2.03% respectively, when using only 50% of the training data. The model also exhibits robustness in the presence of Gaussian noise interference. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper creates a new way to guess where someone is looking called EM-Net. It’s better than other methods because it uses less computer power and still works well. The special trick it uses is called Global Attention Mechanism, which helps it understand how people look at things in general. This makes the guesses more accurate. The paper also tested how well it worked on different pictures of faces and found that it did a better job than other methods even when using less information. |
Keywords
» Artificial intelligence » Attention » Deep learning » Generalization » Machine learning
