# Face Verification on LFW, CFP-FP, AgeDB-30 Aggregate Source: [Face Verification on LFW, CFP-FP, AgeDB-30 Aggregate]() Collection index: [llms.txt]() Data updated: 2026-04-13T09:05:34.258Z Aggregate evaluation of face verification performance across the LFW, CFP-FP, and AgeDB-30 datasets, typically reported using Mean Accuracy. 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 Verification | | Dataset (including any stated split) | LFW, CFP-FP, AgeDB-30 Aggregate | | Metric | Mean Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 9 of 9 | | 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 - [Mean Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Mean Accuracy | Result date | Report paper | | --- | --- | --- | --- | | ResNet200-TopoFR | 99.38 | Not recorded | 69dc78697d0172140ec78986 | | TransFace-L | 99.29 | Not recorded | 69dc78697d0172140ec78986 | | TransFace-B | 99.24 | Not recorded | 69dc78697d0172140ec78986 | | TransFace-S | 99.09 | Not recorded | 69dc78697d0172140ec78986 | | ResNet50-ArcFace | 98.94 | Not recorded | 69dc78697d0172140ec78986 | | FaceLiVTv2-L | 98.69 | Not recorded | 69dc78697d0172140ec78986 | | FaceLiVTv2-M | 98.6 | Not recorded | 69dc78697d0172140ec78986 | | FaceLiVTv2-S | 98.36 | Not recorded | 69dc78697d0172140ec78986 | | FaceLiVTv2-XS | 97.18 | Not recorded | 69dc78697d0172140ec78986 | ### ResNet200-TopoFR Mean Accuracy: 99.38. Result ID: 69dcb1c947f34ce924c3d321. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Year":"'24","Param (M)":"118.8","FLOPs (M)":"23.5G","Training Dataset":"Glint360K","Latency (ms)":"12.24"}. This record does not assert independent reproduction. ### TransFace-L Mean Accuracy: 99.29. Result ID: 69dcb1c947f34ce924c3d322. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Year":"'25","Param (M)":"271.6","FLOPs (M)":"25.4G","Training Dataset":"Glint360K","Latency (ms)":"OOM"}. This record does not assert independent reproduction. ### TransFace-B Mean Accuracy: 99.24. Result ID: 69dcb1c947f34ce924c3d323. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Year":"'25","Param (M)":"124.5","FLOPs (M)":"11.5G","Training Dataset":"Glint360K","Latency (ms)":"18.20"}. This record does not assert independent reproduction. ### TransFace-S Mean Accuracy: 99.09. Result ID: 69dcb1c947f34ce924c3d324. [Reporting paper]() Recorded evaluation configuration: {"Year":"'23","Param (M)":"86.7","FLOPs (M)":"5.8G","Training Dataset":"Glint360K","Latency (ms)":"14.31"}. This record does not assert independent reproduction. ### ResNet50-ArcFace Mean Accuracy: 98.94. Result ID: 69dcb1c947f34ce924c3d325. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Year":"'22","Param (M)":"43.6","FLOPs (M)":"6.3G","Training Dataset":"Glint360K","Latency (ms)":"3.76"}. This record does not assert independent reproduction. ### FaceLiVTv2-L Mean Accuracy: 98.69. Result ID: 69dcb1c947f34ce924c3d326. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param (M)":"8.52","FLOPs (M)":"309","Training Dataset":"Glint360K","Latency (ms)":"0.71(17.2× ↓)"}. This record does not assert independent reproduction. ### FaceLiVTv2-M Mean Accuracy: 98.6. Result ID: 69dcb1c947f34ce924c3d327. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param (M)":"7.02","FLOPs (M)":"258","Training Dataset":"Glint360K","Latency (ms)":"0.65(18.8× ↓)"}. This record does not assert independent reproduction. ### FaceLiVTv2-S Mean Accuracy: 98.36. Result ID: 69dcb1c947f34ce924c3d328. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param (M)":"4.62","FLOPs (M)":"179","Training Dataset":"Glint360K","Latency (ms)":"0.54(22.7× ↓)"}. This record does not assert independent reproduction. ### FaceLiVTv2-XS Mean Accuracy: 97.18. Result ID: 69dcb1c947f34ce924c3d329. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param (M)":"2.9","FLOPs (M)":"90","Training Dataset":"Glint360K","Latency (ms)":"0.43(28.5× ↓)"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.