Summary of Conjnorm: Tractable Density Estimation For Out-of-distribution Detection, by Bo Peng et al.
ConjNorm: Tractable Density Estimation for Out-of-Distribution Detection
by Bo Peng, Yadan Luo, Yonggang Zhang, Yixuan Li, Zhen Fang
First submitted to arxiv on: 27 Feb 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 theoretical framework for post-hoc out-of-distribution (OOD) detection is proposed in this paper, extending distribution considerations to encompass an exponential family of distributions using Bregman divergence. The ConjNorm method is introduced, reframing density function design as a search for the optimal norm coefficient p against the given dataset. An unbiased and analytically tractable estimator of the partition function is devised using Monte Carlo-based importance sampling technique. Experimental results demonstrate that ConjNorm outperforms current best methods on various OOD detection benchmarks by up to 13.25% and 28.19% (FPR95) on CIFAR-100 and ImageNet-1K, respectively. The proposed method’s performance is evaluated on datasets such as CIFAR-100, ImageNet-1K, and others. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper proposes a new way to detect when data is out of the expected range. They use a special type of math called Bregman divergence to design a new method for doing this. This method, called ConjNorm, helps find the right balance between different types of data. The researchers tested their method on several sets of data and found that it performed better than current methods by up to 13% in some cases. |
