# Face Identity Clustering on LFW 12,993 images, 5,749 identities (test)

Source: [Face Identity Clustering on LFW 12,993 images, 5,749 identities (test)](<https://www.sota2.com/research/sota/face-identity-clustering-on-lfw-12-993-images-5-749-identities-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:07:20.580Z

Evaluation of face identity clustering performance using Top-K Recall metrics on the specified test split of the Labeled Faces in the Wild (LFW) dataset.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Face Identity Clustering |
| Dataset (including any stated split) | LFW 12,993 images, 5,749 identities (test) |
| Metric | Top-1 Recall |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Top-1 Recall](<https://www.sota2.com/research/sota/face-identity-clustering-on-lfw-12-993-images-5-749-identities-test.md?metric=Top-1%20Recall>): Higher is better; unit: PERCENT.

- [Top-5 Recall](<https://www.sota2.com/research/sota/face-identity-clustering-on-lfw-12-993-images-5-749-identities-test.md?metric=Top-5%20Recall>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Top-1 Recall | Result date | Report paper |
| --- | --- | --- | --- |
| Ours (Proposed Regression Network) | 16 | Not recorded | 68fa64831e209131cea050ef |
| Tran et al. | 0.1 | Not recorded | 68fa64831e209131cea050ef |
| random | 0.02 | Not recorded | 68fa64831e209131cea050ef |

### Ours (Proposed Regression Network)

Top-1 Recall: 16. Result ID: 698c49ac37894ef7d9cd40bc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa64831e209131cea050ef.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64831e209131cea050ef.md>)

[Implementation](<http://github.com/google/tf_mesh_renderer>)

Recorded evaluation configuration: {"Distance Metric":"VGG-Face (cosine similarity)"}. This record does not assert independent reproduction.

### Tran et al.

Top-1 Recall: 0.1. Result ID: 698c49ac37894ef7d9cd40bb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa64831e209131cea050ef.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63071e209131ce9fc758.md>)

[Implementation](<https://www.sota2.com/www.openu.ac.il/home/hassner/projects/CNN3DMM>)

Recorded evaluation configuration: {"Distance Metric":"VGG-Face (cosine similarity)"}. This record does not assert independent reproduction.

### random

Top-1 Recall: 0.02. Result ID: 698c49ac37894ef7d9cd40ba.

[Reporting paper](<https://www.sota2.com/research/paper/68fa64831e209131cea050ef.md>)

Recorded evaluation configuration: {"Distance Metric":"VGG-Face (cosine similarity)"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.