The following are several research papers about face verification (is this the same person) and recognition (who is this person)
| Title | Accuracy | Approach | Link |
|---|---|---|---|
| FaceNet, by Google | 99.63% | Deep convolutional network with several novel ideas | http://www.cv-foundation.org/openaccess/content_cvpr_2015/papers/Schroff_FaceNet_A_Unified_2015_CVPR_paper.pdf |
| DeepFace, by Facebook | 97.35% | Deep neural network: "employing explicit 3D face modeling in order to apply a piecewise affine transformation, and derive a face representation from a nine-layer deep neural network" | https://www.cs.toronto.edu/~ranzato/publications/taigman_cvpr14.pdf |
| DeepID3 | 99.53% | Ensemble of 2 deep neural networks | https://arxiv.org/pdf/1502.00873v1.pdf |
| GaussianFace | 98.52% | Variant of Gaussian Process | http://luchaochao.me/papers/GaussianFace.pdf |
| Naive-Deep Face Recognition: Touching the Limit of LFW Benchmark or not? | 99.50% | Naive deep convolutional neural network, but uses another database for training (Megvii Face Classification database) | https://arxiv.org/pdf/1501.04690v1.pdf |
| Do We Really Need to Collect Millions of Faces for Effective Face Recognition? | - | - | https://arxiv.org/pdf/1603.07057v2.pdf |