# Emergent Region-Level Facial Correspondence in Frozen Vision Foundation Models

arXiv: https://arxiv.org/abs/2607.14423 · HTML: https://mlsanity.com/paper/2607.14423/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 15 of its 16 references that are in the map.
- First posted on arXiv 15 Jul 2026.

## Builds on (most cited first, 15 of 15)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [Celeb-DF: A Large-scale Challenging Dataset for DeepFake Forensics](https://mlsanity.com/paper/1909.12962/) (2019)
- [Emergent Correspondence from Image Diffusion](https://mlsanity.com/paper/2306.03881/) (2023)
- [Deep ViT Features as Dense Visual Descriptors](https://mlsanity.com/paper/2112.05814/) (2021)
- [A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot Semantic Correspondence](https://mlsanity.com/paper/2305.15347/) (2023)
- [Learning Correspondence from the Cycle-Consistency of Time](https://mlsanity.com/paper/1903.07593/) (2019)
- [TAP-Vid: A Benchmark for Tracking Any Point in a Video](https://mlsanity.com/paper/2211.03726/) (2022)
- [Space-Time Correspondence as a Contrastive Random Walk](https://mlsanity.com/paper/2006.14613/) (2020)
- [General Facial Representation Learning in a Visual-Linguistic Manner](https://mlsanity.com/paper/2112.03109/) (2021)
- [RetinaFace: Single-stage Dense Face Localisation in the Wild](https://mlsanity.com/paper/1905.00641/) (2019)
- [Exploring Temporally-Aware Features for Point Tracking](https://mlsanity.com/paper/2501.12218/) (2025)
- [Semantic Correspondence: Unified Benchmarking and a Strong Baseline](https://mlsanity.com/paper/2505.18060/) (2025)

## Abstract (arXiv)

Frozen self-supervised vision models can align parts of generic objects, but it remains unclear whether this correspondence extends to human faces, where global layout is shared while identity-specific appearance varies sharply. We test whether frozen DINOv3 features define a region-level facial coordinate system: a feature space in which eyes, brows, nose, mouth, skin, and hair remain distinguishable across people and across time without face-specific training. Using DINOv3 ViT-L/16 patch embeddings and FaRL only as a face-part labeling interface, we evaluate cross-identity nearest-neighbor matching and temporal label propagation on 200 CelebDF-v2 real videos. DINOv3 achieves 83.0% region-level semantic accuracy under unconstrained cross-identity matching, compared with a 23.0% area-weighted random baseline, and 95.5% temporal tracking accuracy without a learned temporal module. A no-FaRL control collapses to 0.9%, showing that FaRL supplies semantic initialization while DINOv3 supplies dense spatial correspondence. The strongest correspondence appears at an intermediate layer: block 18 gives a 4.93x same-region versus cross-region discrimination ratio, compared with 1.48x at the final block. Against CLIP ViT-L/14, DINOv3 shows only a small aggregate advantage but a +16.8 pp gain on anatomical regions, indicating that image-level contrastive supervision captures coarse facial layout but not fine-grained anatomical identity. 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.