Image Details

Choose export citation format:

Making Euclid VIS Imaging AI-ready: A Scalable Pipeline for Morphology, Anomaly Detection, and Similarity Search

  • Authors: Jinhui Xie, Yunfei Xu, Zhen Zhang, Lang Chen, Chenzhou Cui

Jinhui Xie et al 2026 The Astrophysical Journal Supplement Series 287 .

  • Provider: AAS Journals

Caption: Figure 2.

Two-dimensional UMAP views of the fixed DINOv2 embedding. Panels show (a) ellipticity, (b) Kron radius, (c) ﹩{\mu }_{{\rm{\max }}}﹩, and (d) ﹩{\mu }_{{\rm{\max }}}-m﹩ for a 120,000-object display sample selected with pandas random_state=42. The saved projection was fitted to the raw embedding with nneighbors = 15, ﹩{\rm{\min }}\_{\rm{dist}}=0.1﹩, Euclidean distance, and UMAP random_state=42. Colors are clipped to the 1st–99th percentiles for visualization. UMAP is used descriptively and does not preserve global distances.

Other Images in This Article
Copyright and Terms & Conditions

Additional terms of reuse