# Deepfake detection on OpenForensics (test) Source: [Deepfake detection on OpenForensics (test)]() Collection index: [llms.txt]() Data updated: 2026-06-11T06:46:17.164Z Evaluation of Deepfake Detection performance on the test split of the OpenForensics dataset, reporting metrics such as Accuracy and AUROC. 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 detection | | Dataset (including any stated split) | OpenForensics (test) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 19 of 19 | | 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 - [Accuracy](): Higher is better; unit: PERCENT. - [AUROC](): Higher is better; unit: PERCENT. - [FPR Delta](): Lower is better; unit: PERCENT. - [TPR Delta](): Higher is better; unit: PERCENT. - [Equal Opportunity](): Lower is better; unit: PERCENT. - [WGA](): Higher is better; unit: PERCENT. - [LD](): Lower is better; unit: SCALAR. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | DeiTFake | 99.22 | Not recorded | 691bfded6750be83fe5ceb21 | | HiFE | 99.03 | Not recorded | 691bfded6750be83fe5ceb21 | | DAW-FDD + FFT | 96.48 | Not recorded | 6a28f33d90966ade919c16fb | | DAW-FDD | 96.33 | Not recorded | 6a28f33d90966ade919c16fb | | DAG-FDD + FFT | 96.32 | Not recorded | 6a28f33d90966ade919c16fb | | PG-FDD + FFT⋆ | 96.3 | Not recorded | 6a28f33d90966ade919c16fb | | DAG-FDD | 95.86 | Not recorded | 6a28f33d90966ade919c16fb | | FFT | 95.81 | Not recorded | 6a28f33d90966ade919c16fb | | DAID + FFT | 95.12 | Not recorded | 6a28f33d90966ade919c16fb | | Tanfoni et al. (2024) | 95.04 | Not recorded | 691bfded6750be83fe5ceb21 | | PG-FDD | 94.86 | Not recorded | 6a28f33d90966ade919c16fb | | DAID | 94.42 | Not recorded | 6a28f33d90966ade919c16fb | | FF Discover | 93.62 | Not recorded | 6a28f33d90966ade919c16fb | | Fair-FLIP | 92.95 | Not recorded | 6a28f33d90966ade919c16fb | | BASE | 92.88 | Not recorded | 6a28f33d90966ade919c16fb | | BPFA | 92.87 | Not recorded | 6a28f33d90966ade919c16fb | | FILTER | 92.04 | Not recorded | 691bfded6750be83fe5ceb21 | | FF Max | 91.86 | Not recorded | 6a28f33d90966ade919c16fb | | COMICS | 88.2 | Not recorded | 691bfded6750be83fe5ceb21 | ### DeiTFake Accuracy: 99.22. Result ID: 69aa95c06dbb7075b954742e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Dual-Stage Trained DeiT Model with Complex Augmentations"}. This record does not assert independent reproduction. ### HiFE Accuracy: 99.03. Result ID: 69aa95c06dbb7075b954742d. [Reporting paper]() Recorded evaluation configuration: {"Methodology":"High-frequency enhancement network for compressed content"}. This record does not assert independent reproduction. ### DAW-FDD + FFT Accuracy: 96.48. Result ID: 6a2a59b4fbc373592c26d190. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### DAW-FDD Accuracy: 96.33. Result ID: 6a2a59b4fbc373592c26d18c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### DAG-FDD + FFT Accuracy: 96.32. Result ID: 6a2a59b4fbc373592c26d18f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### PG-FDD + FFT⋆ Accuracy: 96.3. Result ID: 6a2a59b4fbc373592c26d191. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### DAG-FDD Accuracy: 95.86. Result ID: 6a2a59b4fbc373592c26d18b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### FFT Accuracy: 95.81. Result ID: 6a2a59b4fbc373592c26d188. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"Post-Processing Methods"}. This record does not assert independent reproduction. ### DAID + FFT Accuracy: 95.12. Result ID: 6a2a59b4fbc373592c26d192. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### Tanfoni et al. (2024) Accuracy: 95.04. Result ID: 69aa95c06dbb7075b954742c. [Reporting paper]() Recorded evaluation configuration: {"Methodology":"Facial Segmentation + transfer learning (DeepLabV3+)"}. This record does not assert independent reproduction. ### PG-FDD Accuracy: 94.86. Result ID: 6a2a59b4fbc373592c26d18d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### DAID Accuracy: 94.42. Result ID: 6a2a59b4fbc373592c26d18e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"In-Processing Methods"}. This record does not assert independent reproduction. ### FF Discover Accuracy: 93.62. Result ID: 6a2a59b4fbc373592c26d18a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"Post-Processing Methods"}. This record does not assert independent reproduction. ### Fair-FLIP Accuracy: 92.95. Result ID: 6a2a59b4fbc373592c26d186. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"Post-Processing Methods"}. This record does not assert independent reproduction. ### BASE Accuracy: 92.88. Result ID: 6a2a59b4fbc373592c26d193. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"REFERENCE"}. This record does not assert independent reproduction. ### BPFA Accuracy: 92.87. Result ID: 6a2a59b4fbc373592c26d187. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"Post-Processing Methods"}. This record does not assert independent reproduction. ### FILTER Accuracy: 92.04. Result ID: 69aa95c06dbb7075b954742b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Multi-face relationship learning + global feature aggregation"}. This record does not assert independent reproduction. ### FF Max Accuracy: 91.86. Result ID: 6a2a59b4fbc373592c26d189. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"MobileNetV3-Small","Category":"Post-Processing Methods"}. This record does not assert independent reproduction. ### COMICS Accuracy: 88.2. Result ID: 69aa95c06dbb7075b954742a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Bi-grained contrastive learning with integrated face extraction"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.