Author

Izaldein Al-Zyoud

h-index 1 View on Semantic Scholar

We have 1 of 4 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Jul 2026

Emergent Region-Level Facial Correspondence in Frozen Vision Foundation Models

These results establish frozen DINOv3 as a strong zero-shot representation for region-level facial correspondence and identify intermediate self-supervised features as the most useful layer for dense face analysis.

Izaldein Al-Zyoud, Abdulmotaleb El Saddik · arXiv.org · 0 citations