# Facial Expression Recognition on SFEW 2.0 (val) Source: [Facial Expression Recognition on SFEW 2.0 (val)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:28:38.719Z Evaluation of Facial Expression Recognition performance on the validation split of the SFEW 2.0 dataset, typically measured using 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 | Facial Expression Recognition | | Dataset (including any stated split) | SFEW 2.0 (val) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 31 of 31 | | 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. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | APVIT | 61.92 | Not recorded | 68fa6e301e209131cea3e6f7 | | MA-Net | 59.4 | Not recorded | 68fa6e301e209131cea3e6f7 | | AHA | 58.89 | Not recorded | 68fa6e301e209131cea3e6f7 | | Model-4 | 58.14 | Not recorded | 68fa64631e209131cea0446f | | DAN | 57.88 | Not recorded | 68fa6e301e209131cea3e6f7 | | APM | 57.57 | Not recorded | 68fa6e301e209131cea3e6f7 | | PAT-Res101 + attr | 57.57 | Not recorded | 68fa6e301e209131cea3e6f7 | | Model-3 | 57.48 | Not recorded | 68fa64631e209131cea0446f | | Model-2 | 56.72 | Not recorded | 68fa64631e209131cea0446f | | RAN | 56.4 | Not recorded | 68fa6e301e209131cea3e6f7 | | CMU†† | 55.96 | Not recorded | 68fa62d41e209131ce9fb6df | | EmotiW-1 | 55.96 | Not recorded | 68fa64631e209131cea0446f | | Model-1 | 55.4 | Not recorded | 68fa64631e209131cea0446f | | LBAN-IL | 55.28 | Not recorded | 68fa6e301e209131cea3e6f7 | | VGG-VD-16 | 54.82 | Not recorded | 68fa62d41e209131ce9fb6df | | VGG-VD-16 | 54.82 | Not recorded | 68fa64631e209131cea0446f | | Baseline Model | 54.45 | Not recorded | 68fa64631e209131cea0446f | | PAT-Res101 | 53.9 | Not recorded | 68fa6e301e209131cea3e6f7 | | HDC†† | 52.8 | Not recorded | 68fa62d41e209131ce9fb6df | | EmotiW-2 | 52.8 | Not recorded | 68fa64631e209131cea0446f | | HDC\* | 52.5 | Not recorded | 68fa62d41e209131ce9fb6df | | Kim's CNN | 52.5 | Not recorded | 68fa6e301e209131cea3e6f7 | | CMU\* | 52.29 | Not recorded | 68fa62d41e209131ce9fb6df | | DLP-CNN | 51.05 | Not recorded | 68fa6e301e209131cea3e6f7 | | Resnet-50 | 50.91 | Not recorded | 68fa62d41e209131ce9fb6df | | Resnet-50 | 47.48 | Not recorded | 68fa62d41e209131ce9fb6df | | VGG-M | 47.02 | Not recorded | 68fa62d41e209131ce9fb6df | | VGG-VD-16 | 43.58 | Not recorded | 68fa62d41e209131ce9fb6df | | VGG-M | 42.9 | Not recorded | 68fa62d41e209131ce9fb6df | | AlexNet | 38.07 | Not recorded | 68fa62d41e209131ce9fb6df | | AlexNet | 37.67 | Not recorded | 68fa62d41e209131ce9fb6df | ### APVIT Accuracy: 61.92. Result ID: 699009013a6c314849fda38c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Pre-training":"RAF-DB"}. This record does not assert independent reproduction. ### MA-Net Accuracy: 59.4. Result ID: 699009013a6c314849fda389. [Reporting paper]() Recorded evaluation configuration: {"Pre-training":"FER2013"}. This record does not assert independent reproduction. ### AHA Accuracy: 58.89. Result ID: 699009013a6c314849fda388. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Model-4 Accuracy: 58.14. Result ID: 698c2f40112aeb9460f00b48. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Covariance Pooling":"true","SPD Layers":"Cov-BiRe-LogEig","Architecture":"FC2000-FC512-FC7"}. This record does not assert independent reproduction. ### DAN Accuracy: 57.88. Result ID: 699009013a6c314849fda387. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### APM Accuracy: 57.57. Result ID: 699009013a6c314849fda386. [Reporting paper]() Recorded evaluation configuration: {"Pre-training":"RAF-DB"}. This record does not assert independent reproduction. ### PAT-Res101 + attr Accuracy: 57.57. Result ID: 699009013a6c314849fda38b. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ResNet-101","Features":"Attributes included"}. This record does not assert independent reproduction. ### Model-3 Accuracy: 57.48. Result ID: 698c2f40112aeb9460f00b47. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Covariance Pooling":"true","SPD Layers":"Cov-BiRe-BiRe-LogEig"}. This record does not assert independent reproduction. ### Model-2 Accuracy: 56.72. Result ID: 698c2f40112aeb9460f00b46. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Covariance Pooling":"true","SPD Layers":"Cov-BiRe-LogEig","Architecture":"FC2000-FC128-FC7"}. This record does not assert independent reproduction. ### RAN Accuracy: 56.4. Result ID: 699009013a6c314849fda385. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"VGG16+ResNet18"}. This record does not assert independent reproduction. ### CMU†† Accuracy: 55.96. Result ID: 698be297ecace47ef55991cf. [Reporting paper]() Recorded evaluation configuration: {"Pretraining":"FER combined","ensemble":true}. This record does not assert independent reproduction. ### EmotiW-1 Accuracy: 55.96. Result ID: 698c2f40112aeb9460f00b4a. [Reporting paper]() Recorded evaluation configuration: {"year":"2015"}. This record does not assert independent reproduction. ### Model-1 Accuracy: 55.4. Result ID: 698c2f40112aeb9460f00b45. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Covariance Pooling":"true","SPD Layers":"Cov-BiRe-LogEig"}. This record does not assert independent reproduction. ### LBAN-IL Accuracy: 55.28. Result ID: 699009013a6c314849fda384. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### VGG-VD-16 Accuracy: 54.82. Result ID: 698be297ecace47ef55991cc. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces+FER"}. This record does not assert independent reproduction. ### VGG-VD-16 Accuracy: 54.82. Result ID: 698c2f40112aeb9460f00b49. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Baseline Model Accuracy: 54.45. Result ID: 698c2f40112aeb9460f00b44. [Reporting paper]() Recorded evaluation configuration: {"Covariance Pooling":"false"}. This record does not assert independent reproduction. ### PAT-Res101 Accuracy: 53.9. Result ID: 699009013a6c314849fda38a. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ResNet-101"}. This record does not assert independent reproduction. ### HDC†† Accuracy: 52.8. Result ID: 698be297ecace47ef55991d0. [Reporting paper]() Recorded evaluation configuration: {"Pretraining":"FER + TFD","ensemble":true}. This record does not assert independent reproduction. ### EmotiW-2 Accuracy: 52.8. Result ID: 698c2f40112aeb9460f00b4b. [Reporting paper]() Recorded evaluation configuration: {"year":"2015"}. This record does not assert independent reproduction. ### HDC\* Accuracy: 52.5. Result ID: 698be297ecace47ef55991ce. [Reporting paper]() Recorded evaluation configuration: {"Pretraining":"FER + TFD","best\_single\_CNN":true}. This record does not assert independent reproduction. ### Kim's CNN Accuracy: 52.5. Result ID: 699009013a6c314849fda383. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### CMU\* Accuracy: 52.29. Result ID: 698be297ecace47ef55991cd. [Reporting paper]() Recorded evaluation configuration: {"Pretraining":"FER combined","best\_single\_CNN":true}. This record does not assert independent reproduction. ### DLP-CNN Accuracy: 51.05. Result ID: 699009013a6c314849fda382. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Resnet-50 Accuracy: 50.91. Result ID: 698be297ecace47ef55991cb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces+FER"}. This record does not assert independent reproduction. ### Resnet-50 Accuracy: 47.48. Result ID: 698be297ecace47ef55991c7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces"}. This record does not assert independent reproduction. ### VGG-M Accuracy: 47.02. Result ID: 698be297ecace47ef55991ca. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces+FER"}. This record does not assert independent reproduction. ### VGG-VD-16 Accuracy: 43.58. Result ID: 698be297ecace47ef55991c8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces"}. This record does not assert independent reproduction. ### VGG-M Accuracy: 42.9. Result ID: 698be297ecace47ef55991c6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces"}. This record does not assert independent reproduction. ### AlexNet Accuracy: 38.07. Result ID: 698be297ecace47ef55991c9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces+FER"}. This record does not assert independent reproduction. ### AlexNet Accuracy: 37.67. Result ID: 698be297ecace47ef55991c5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretraining":"VGGFaces"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.