# Deepfake Attribution on OW-DFA-40 Protocol-1

Source: [Deepfake Attribution on OW-DFA-40 Protocol-1](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1>)

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

Data updated: 2026-02-26T05:34:52.209Z

Evaluation of Deepfake Attribution performance on the OW-DFA-40 Protocol-1 dataset, reporting metrics related to accuracy, normalized mutual information (NMI), and adjusted rand index (ARI) across known and new classes.

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 | Deepfake Attribution |
| Dataset (including any stated split) | OW-DFA-40 Protocol-1 |
| Metric | All Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [All Accuracy](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1.md?metric=All%20Accuracy>): Higher is better; unit: PERCENT.

- [All NMI](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1.md?metric=All%20NMI>): Higher is better; unit: PERCENT.

- [All ARI](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1.md?metric=All%20ARI>): Higher is better; unit: PERCENT.

- [New Accuracy](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1.md?metric=New%20Accuracy>): Higher is better; unit: PERCENT.

- [New NMI](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1.md?metric=New%20NMI>): Higher is better; unit: PERCENT.

- [New ARI](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1.md?metric=New%20ARI>): Higher is better; unit: PERCENT.

- [Known Accuracy](<https://www.sota2.com/research/sota/deepfake-attribution-on-ow-dfa-40-protocol-1.md?metric=Known%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | All Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| CAL | 88.3 | Not recorded | 6940e77e456146936284f6ed |
| CAL (w/o KU) | 87.5 | Not recorded | 6940e77e456146936284f6ed |
| CDAL | 84.3 | Not recorded | 6940e77e456146936284f6ed |
| CPL | 82.6 | Not recorded | 6940e77e456146936284f6ed |
| PALGCD | 76.5 | Not recorded | 6940e77e456146936284f6ed |
| LPS | 76.1 | Not recorded | 6940e77e456146936284f6ed |
| OwMatch | 74.1 | Not recorded | 6940e77e456146936284f6ed |
| SimGCD | 73.3 | Not recorded | 6940e77e456146936284f6ed |
| LegoGCD | 70.9 | Not recorded | 6940e77e456146936284f6ed |
| ProtoGCD | 62.9 | Not recorded | 6940e77e456146936284f6ed |

### CAL

All Accuracy: 88.3. Result ID: 69984df95698665ec4ab6d07.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### CAL (w/o KU)

All Accuracy: 87.5. Result ID: 69984df95698665ec4ab6d06.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"unknown"}. This record does not assert independent reproduction.

### CDAL

All Accuracy: 84.3. Result ID: 69984df95698665ec4ab6d05.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b01e209131cea7ee46.md>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### CPL

All Accuracy: 82.6. Result ID: 69984df95698665ec4ab6d04.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70ab1e209131cea4c579.md>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### PALGCD

All Accuracy: 76.5. Result ID: 69984df95698665ec4ab6d03.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76ee1e209131cea70076.md>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### LPS

All Accuracy: 76.1. Result ID: 69984df95698665ec4ab6d00.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f41e209131cea59863.md>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### OwMatch

All Accuracy: 74.1. Result ID: 69984df95698665ec4ab6cff.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75851e209131cea68592.md>)

[Implementation](<https://github.com/niusj03/OwMatch>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### SimGCD

All Accuracy: 73.3. Result ID: 69984df95698665ec4ab6cfe.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71741e209131cea51542.md>)

[Implementation](<https://github.com/CVMI-Lab/SimGCD>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### LegoGCD

All Accuracy: 70.9. Result ID: 69984df95698665ec4ab6d01.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77481e209131cea723c4.md>)

[Implementation](<https://github.com/Cliffia123/LegoGCD>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

### ProtoGCD

All Accuracy: 62.9. Result ID: 69984df95698665ec4ab6d02.

[Reporting paper](<https://www.sota2.com/research/paper/6940e77e456146936284f6ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77e21e209131cea75cd4.md>)

Recorded evaluation configuration: {"Knowledge of number of novel forgery types (KU)":"known"}. This record does not assert independent reproduction.

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