用扩散模型估算射电星系图像的内在维度，发现异常样本维度更高。
Intrinsic Dimension Estimation for Radio Galaxy Zoo using Diffusion Models
- 基于得分函数的扩散模型估计数据内在维度
- 射电源异常度越高，内在维度越大，整体高于自然图像数据集
- 适合研究自监督学习表征质量的科研人员参考
本文使用基于得分的扩散模型估计射电星系动物园（Radio Galaxy Zoo, RGZ）数据集的内在维度（iD）。通过分析贝叶斯神经网络（BNN）能量分数的变化，考察其与数据相似性（相对于MiraBest子集）的关系。结果表明，分布外（out-of-distribution）样本的iD值更高，且整个RGZ数据集的iD显著高于典型自然图像数据集。进一步分析显示，不同法纳罗夫-瑞利（Fanaroff-Riley, FR）形态类别间无明显差异，但信号噪声比（SNR）越低，iD越小，呈现弱趋势。未来工作可利用iD与能量分数的关系，定量评估和改进自监督学习算法的表征能力。
原文摘要 · Abstract (English)
In this work, we estimate the intrinsic dimension (iD) of the Radio Galaxy Zoo (RGZ) dataset using a score-based diffusion model. We examine how the iD estimates vary as a function of Bayesian neural network (BNN) energy scores, which measure how similar the radio sources are to the MiraBest subset of the RGZ dataset. We find that out-of-distribution sources exhibit higher iD values, and that the overall iD for RGZ exceeds those typically reported for natural image datasets. Furthermore, we analyse how iD varies across Fanaroff-Riley (FR) morphological classes and as a function of the signal-to-noise ratio (SNR). While no relationship is found between FR I and FR II classes, a weak trend toward higher SNR at lower iD. Future work using the RGZ dataset could make use of the relationship between iD and energy scores to quantitatively study and improve the representations learned by various self-supervised learning algorithms.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。