Summary of Energy-based Hopfield Boosting For Out-of-distribution Detection, by Claus Hofmann et al.
Energy-based Hopfield Boosting for Out-of-Distribution Detection
by Claus Hofmann, Simon Schmid, Bernhard Lehner, Daniel Klotz, Sepp Hochreiter
First submitted to arxiv on: 14 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV)
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 Hopfield Boosting approach enhances out-of-distribution (OOD) detection by leveraging modern Hopfield energy (MHE) to sharpen the decision boundary between in-distribution and OOD data. This boosting method focuses on hard-to-distinguish auxiliary outlier examples that lie close to this boundary, improving OOD detection performance compared to traditional methods. The authors achieve a new state-of-the-art in OOD detection with outlier exposure, achieving significant improvements in FPR95 metrics on CIFAR-10 (2.28 → 0.92) and CIFAR-100 (11.76 → 7.94). |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Hopfield Boosting is a way to improve out-of-distribution detection by using extra information. This helps models be more accurate when they’re not sure about something. The approach uses a special kind of energy called modern Hopfield energy to make the model better at telling apart normal data and unusual data. This makes it good at detecting things that are different from what it’s used to. The results show that this method is much better than others at doing this, especially on certain types of images. |
Keywords
» Artificial intelligence » Boosting
