FaceNet: A unified embedding for face recognition and clustering
Open accessThe DOI matches its Crossref record; 23 of the 23 references it cites were found in a registry or an index.
Abstract
Despite significant recent advances in the field of face recognition [10, 14, 15, 17], implementing face verification and recognition efficiently at scale presents serious challenges to current approaches. In this paper we present a system, called FaceNet, that directly learns a mapping from face images to a compact Euclidean space where distances directly correspond to a measure offace similarity. Once this space has been produced, tasks such as face recognition, verification and clustering can be easily implemented using standard techniques with FaceNet embeddings asfeature vectors. Our method uses a deep convolutional network trained to directly optimize the embedding itself, rather than an intermediate bottleneck layer as in previous deep learning approaches. To train, we use triplets of roughly aligned matching / non-matching face patches generated using a novel online triplet mining method. The benefit of our approach is much greater representational efficiency: we achieve state-of-the-artface recognition performance using only 128-bytes perface. On the widely used Labeled Faces in the Wild (LFW) dataset, our system achieves a new record accuracy of 99.63%. On YouTube Faces DB it achieves 95.12%. Our system cuts the error rate in comparison to the best published result [15] by 30% on both datasets.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Current face recognition systems [8,24,43], already reach very high levels of accuracy in public non-forensic benchmarks, and it is expected that in the coming years they will keep improving.”
From Calibration of score based likelihood ratio estimation in automated forensic facial image comparison (2022) · cites this paper as [43]
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“An increasing number of UDA techniques have been applied in HAR; however, most of them work between domains that share the same feature space; for example, transferring the learn models between users [44] or between sensor wearing positions [32].”
From ContrasGAN: Unsupervised domain adaptation in Human Activity Recognition via adversarial and contrastive learning (2021) · cites this paper as [32]
Experimental — accuracy not yet measuredIn the citing paper
“To address this issue, a representation learning method using triplet loss is proposed to directly learn discriminative and compact face embedding [99].”
From Empowering things with intelligence: a survey of the progress, challenges, and opportunities in artificial intelligence of things (2021) · cites this paper as [99]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562 · cited 8×
“We would like to thank Johannes Steffens for his discussions and great insights on face recognition and Christian Szegedy for providing new network architectures like [16] and discussing network design choices.”
MethodResultInfluentialLearning representations by back-propagating errors
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986) · Nature · doi:10.1038/323533a0 · Cited by 31,892 · cited 2×
“In all our experiments we train the CNN using Stochastic Gradient Descent (SGD) with standard backprop [8, 11] and AdaGrad [5].”
MethodVisualizing and understanding convolutional networks
Zeiler, M. D., & Fergus, R. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10590-1_53 · Cited by 15,839 · cited 5×
“This table show the structure of our Zeiler&Fergus [22] based model with 1×1 convolutions inspired by [9].”
BackgroundMethodInfluentialBackpropagation Applied to Handwritten Zip Code Recognition
LeCun, Y., Boser, B., Denker, J. S., Henderson, D., Howard, R. E., Hubbard, W., & Jackel, L. D. (1989) · Neural Computation · doi:10.1162/neco.1989.1.4.541 · Cited by 11,891 · cited 2×
“In all our experiments we train the CNN using Stochastic Gradient Descent (SGD) with standard backprop [8, 11] and AdaGrad [5].”
Method(2010) · Cited by 8,261
DeepFace: Closing the Gap to Human-Level Performance in Face Verification
Taigman, Y., Yang, M., Ranzato, M., & Wolf, L. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.220 · Cited by 6,542 · cited 9×
“[17], for example, performs a complex 3D alignment.”
BackgroundMethodInfluentialBengio, Y., Louradour, J., Collobert, R., & Weston, J. (2009) · Proceedings of the 26th Annual International Conference on Machine Learning · doi:10.1145/1553374.1553380 · Cited by 5,366 · cited 1×
“Choosing which triplets to use turns out to be very important for achieving good performance and, inspired by curriculum learning [1], we present a novel online negative exemplar mining strategy which ensures consistently increasing difficulty of triplets as the network trains.”
BackgroundLarge Scale Distributed Deep Networks
Dean, J. B., Corrado, G. S., Monga, R., Chen, K., Devin, M., Mao, M., Ranzato, M., Senior, A., Tucker, P. A., Yang, K., et al. (2012) · neural information processing systems · Cited by 4,489
Distance Metric Learning for Large Margin Nearest Neighbor Classification
Weinberger, K. Q., & Saul, L. K. (2009) · Journal of Machine Learning Research · doi:10.5555/1577069.1577078 · Cited by 4,044 · cited 4×
“D embedding using a tripletbased loss function based on LMNN [19].”
MethodInfluentialDeep Learning Face Representation by Joint Identification-Verification
Sun, Y., Wang, X., & Tang, X. (2014) · arXiv · doi:10.48550/arxiv.1406.4773 · Cited by 1,906 · cited 4×
“the one using pairs of positives and negatives, as used in [14] Eq.”
BackgroundMethodInfluential
Research works printed but not linked: 2
- huang, 2007, Labeled faces in the wild: A database for studying face recognition in unconstrained environments, Technical Report, 7
- lin, 2013, Network in network, CaRR abs/1312 4400
Cited by
Items 801–900 of 1,094
Joint Face Alignment and 3D Face Reconstruction with Application to Face Recognition
Liu, F., Zhao, Q., Liu, X., & Zeng, D. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2018.2885995 · Cited by 133
Associative Deep Clustering: Training a Classification Network with No Labels
Haeusser, P., Plapp, J., Golkov, V., Aljalbout, E., & Cremers, D. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-12939-2_2 · Cited by 135
Deep Randomized Ensembles for Metric Learning
Xuan, H., Souvenir, R., & Pless, R. B. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01270-0_44 · Cited by 113
Benchmarking of deep learning irradiance forecasting models from sky images – An in-depth analysis
Paletta, Q., Arbod, G., & Lasenby, J. (2021) · Solar Energy · doi:10.1016/j.solener.2021.05.056 · Cited by 142
Spatial-Channel Enhanced Transformer for Visible-Infrared Person Re-Identification
Zhao, J., Wang, H., Zhou, Y. Z., Yao, R., Chen, S., & Saddik, A. E. (2022) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2022.3163847 · Cited by 124
Contrastive self-supervised learning: review, progress, challenges and future research directions
Kumar, P., Rawat, P., & Chauhan, S. (2022) · International Journal of Multimedia Information Retrieval · doi:10.1007/s13735-022-00245-6 · Cited by 140
Image denoising in the deep learning era
Izadi, S., Sutton, D. B., & Hamarneh, G. (2022) · Artificial Intelligence Review · doi:10.1007/s10462-022-10305-2 · Cited by 150
“The noise generative and camera encoding networks are trained jointly along with a discriminator supervised by adversarial, feature matching (Wang et al. 2018) and triplet losses (Schroff et al. 2015).”
MethodAudio self-supervised learning: A survey
Liu, S., Mallol-Ragolta, A., Parada‐Cabaleiro, E., Qian, K., Jing, X., Kathan, A., Hu, B., & Schuller, B. W. (2022) · Patterns · doi:10.1016/j.patter.2022.100616 · Cited by 189
FReeNet: Multi-Identity Face Reenactment
Zhang, J., Zeng, X., Wang, M., Pan, Y., Liu, L., Liu, Y., Ding, Y., & Fan, C. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00537 · Cited by 111
“Furthermore, we combine the triplet loss [28] and the perceptual loss [14] to form a novel triplet perceptual loss, which helps generate more detailed images as well as decouple the appearance and geometry information.”
MethodPersonalized Age Progression with Bi-Level Aging Dictionary Learning
Shu, X., Tang, J., Li, Z., Lai, H., Zhang, L., & Yan, S. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2705122 · Cited by 95
Face Clustering: Representation and Pairwise Constraints
Shi, Y., Otto, C., & Jain, A. K. (2018) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2018.2796999 · Cited by 127
“[25], in their FaceNet work, abandoned the classification layer and instead introduced the triplet loss to directly learn an embedding space where feature vectors of different identities could be separated with Euclidean distance.”
BackgroundDeep learning and face recognition: the state of the art
Balaban, S. (2015) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2181526 · Cited by 152
Mining on Manifolds: Metric Learning Without Labels
İşcen, A., Tolias, G., Avrithis, Y., & Chum, O. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00797 · Cited by 142
Yang, H., Zhang, Z., & Yin, L. (2018) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg.2018.00050 · Cited by 124
Deep Variational Metric Learning
Lin, X., Duan, Y., Dong, Q., Lu, J., & Zhou, J. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01267-0_42 · Cited by 133
Coupled Deep Learning for Heterogeneous Face Recognition
Wu, X., Song, L., He, R., & Tan, T. N. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.11500 · Cited by 119
A Computer-Vision Based Application for Student Behavior Monitoring in Classroom
Anh, B. N., Son, N. T., Lam, P. T., Chi, L. P., Tuan, N. H., Dat, N. C., Trung, N. H., Aftab, M. U., & Dinh, T. V. (2019) · Applied Sciences · doi:10.3390/app9224729 · Cited by 130
“There are some popular, recently-developed face embedding algorithms, such as DeepFace [36], FaceNet [37], VGGFace [38], SphereFace [39], and the state-of-the-art ArcFace [40] which scores 99.83% accuracy on Labeled Faces in the Wild Home (LFW) dataset.”
MethodInfluentialFace Spoofing Detection Based on Local Ternary Label Supervision in Fully Convolutional Networks
Sun, W., Song, Y., Chen, C., Huang, J., & Kot, A. C. (2020) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2020.2985530 · Cited by 111
“Deep Convolutional Neural Network (CNN)-based face ver-ification algorithms [2], [3] have achieved great success in recent years with the development of large training datasets and Graphics Processing Unit (GPU) hardware architecture.”
MethodDave, S., Baghdadi, R., Nowatzki, T., Avancha, S., Shrivastava, A., & Li, B. (2021) · Proceedings of the IEEE · doi:10.1109/jproc.2021.3098483 · Cited by 111
Online Knowledge Distillation via Mutual Contrastive Learning for Visual Recognition
Yang, C., An, Z., Zhou, H., Zhuang, F., Xu, Y., & Zhang, Q. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3257878 · Cited by 150
“Contrastive learning is widely used for deep metric learning, such as image classification [24], face recognition [23], person re-identification [46] and image retrieval [47].”
BackgroundMARLIN: Masked Autoencoder for facial video Representation LearnINg
Cai, Z., Ghosh, S., Stefanov, K., Dhall, A., Cai, J., Rezatofighi, H., Haffari, R., & Hayat, M. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.00150 · Cited by 109
Learning towards Minimum Hyperspherical EnergyPreprint
Liu, W., Lin, R., Liu, Z., Liu, L., Yu, Z., Dai, B., & Song, L. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.09298 · Cited by 156
“22 Triplet Loss [41] 98.”
Revisiting Training Strategies and Generalization Performance in Deep Metric LearningPreprint
Roth, K., Milbich, T., Sinha, S., Gupta, P., Ommer, B., & Cohen, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.08473 · Cited by 201
“…become increasingly important for applications ranging from image retrieval (Movshovitz-Attias et al., 2017; Roth et al., 2019; Wu et al., 2017; Lin et al., 2018) to zero-shot classification (Schroff et al., 2015; Sanakoyeu et al., 2019) and face verification (Hu et al., 2014; Liu et al., 2017).”
BackgroundMethodResultInfluentialAn Adversarial Approach to Hard Triplet Generation
Zhao, Y., Jin, Z., Qi, G.-J., Lu, H., & Hua, X. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01240-3_31 · Cited by 116
“Deep metric learning is of great practical importance and has shown promising results in many tasks, such as image retrieval [20, 32, 44, 31], face recognition [26, 34, 40], person re-identification [1, 28, 46], etc.”
BackgroundCOCAS: A Large-Scale Clothes Changing Person Dataset for Re-Identification
Yu, S., Li, S., Chen, D., Zhao, R., Yan, J., & Qiao, Y. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00346 · Cited by 124
Zhao, C., & Shen, W. (2023) · Advanced Engineering Informatics · doi:10.1016/j.aei.2023.102262 · Cited by 119
AdamP: Slowing Down the Slowdown for Momentum Optimizers on Scale-invariant WeightsPreprint
Heo, B., Chun, S., Oh, S. J., Han, D., Yun, S., Kim, G., Uh, Y., & Ha, J.-W. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.08217 · Cited by 212
VIPLFaceNet: an open source deep face recognition SDK
Liu, X., Kan, M., Wu, W., Shan, S., & Chen, X. (2017) · Frontiers of Computer Science · doi:10.1007/s11704-016-6076-3 · Cited by 119
“Representative methods including LDML-MkNN [39], Multishot [40], LE [12], AssociatePredict [41], Tom-vs-Pete [14], Fisher Vector Face [22], High-dim LBP [4], TL Joint Bayesian [42], DeepFace [15], DeepID [16], DeepID2 [17], Gaussian Face [43], VGGFace [44], DeepID2+ [18], and FaceNet [19] are shown.”
MethodDeep Multi-View Feature Learning for Person Re-Identification
Tao, D., Guo, Y., Yu, B., Pang, J., & Yu, Z. (2018) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2017.2726580 · Cited by 109
“Given a CNN, and treating it as an end-to-end system, triplet loss [46] can be directly related to the practical problems of person re-identification; that is, triplet loss training aims to learn an optimal feature representation that maximizes the relative distance.”
BackgroundCooperative Training of Deep Aggregation Networks for RGB-D Action Recognition
Wang, P., Li, W., Wan, J., Ogunbona, P., & Liu, X. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.12228 · Cited by 107
A Haze Prediction Method Based on One-Dimensional Convolutional Neural Network
Zhang, Z., Tian, J., Huang, W., Yin, L., Zheng, W., & Liu, S. (2021) · Atmosphere · doi:10.3390/atmos12101327 · Cited by 111
A union of deep learning and swarm-based optimization for 3D human action recognition
Basak, H., Kundu, R., Singh, P. K., Ijaz, M. F., Woźniak, M., & Sarkar, R. (2022) · Scientific Reports · doi:10.1038/s41598-022-09293-8 · Cited by 119
“The Inception-ResNet network has been previously successfully applied on tasks like face recognition 47 .”
MethodSpoofing Deep Face Recognition with Custom Silicone Masks
Bhattacharjee, S., Mohammadi, A., & Marcel, S. (2018) · IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS) · doi:10.1109/btas.2018.8698550 · Cited by 110
“[18] report an accuracy of 99.”
Swire: Sketch-based User Interface Retrieval
Huang, F., Canny, J., & Nichols, J. W. (2019) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290605.3300334 · Cited by 151
“The model is trained with a Triplet Loss function [23, 28] that involves the neural-network outputs of three inputs: an ‘anchor’ sketch s , a ‘positive’ matching screenshot i and a ′ ‘negative’ mismatched screenshot i .”
MethodRobust Multimodal Emotion Recognition from Conversation with Transformer-Based Crossmodality Fusion
Xie, B., Sidulova, M., & Park, C. H. (2021) · Sensors · doi:10.3390/s21144913 · Cited by 108
“FaceNet is a deep CNN model that has been utilized to identify facial features from the image inputs [53].”
MethodDG-Softmax: A new domain generalization intelligent fault diagnosis method for planetary gearboxes
Qian, Q., Wen, Q., Tang, R., & Qin, Y. (2025) · Reliability Engineering & System Safety · doi:10.1016/j.ress.2025.111057 · Cited by 119
Rethinking Feature Discrimination and Polymerization for Large-scale RecognitionPreprint
Liu, Y., Li, H., & Wang, X. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1710.00870 · Cited by 141
“The triplet loss [33] has a constraint on feature embedding ( (cid:107) f (cid:107) 2 = 1 ) and thus the learned features are scatted on a hypersphere.”
BackgroundMethodInfluentialRethinking Softmax Cross-Entropy Loss for Adversarial Robustness
Pang, T., Xu, K., Dong, Y., Du, C., Chen, N., & Zhu, J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.10626 · Cited by 91
Learning Deep Off-the-Person Heart Biometrics Representations
Luz, E. J. D. S., Moreira, G. J. P., Oliveira, L. S., Schwartz, W. R., & Menotti, D. (2017) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2017.2784362 · Cited by 107
Deep Learning and Big Data in Healthcare: A Double Review for Critical Beginners
Bote-Curiel, L., Muñoz-Romero, S., Guerrero-Curieses, A., & Rojo‐Álvarez, J. L. (2019) · Applied Sciences · doi:10.3390/app9112331 · Cited by 129
“Face recognition is one of the hottest computer vision applications with great commercial interest [67], like Google’s FaceNet [68] and Facebook’s DeepFace [69].”
BackgroundZhao, J., Cheng, Y., Cheng, Y., Yang, Y., Zhao, F., Li, J., Liu, H., Yan, S., & Feng, J. (2019) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v33i01.33019251 · Cited by 104
RSPNet: Relative Speed Perception for Unsupervised Video Representation Learning
Chen, P., Huang, D. Y., He, D., Long, X., Zeng, R., Wen, S., Tan, M., & Gan, C. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i2.16189 · Cited by 140
Jin, B., Cruz, L., & Gonçalves, N. (2022) · IEEE Sensors Journal · doi:10.1109/jsen.2022.3197235 · Cited by 113
“In this section, classic ML and DL models, including PCA [10], ICA [11], FaceNet [17], and InsightFace [19], are selected as face recognition methods.”
MethodPeng, L., Jian, S., Kan, Z., Qiao, L., & Li, D. (2023) · Information Processing & Management · doi:10.1016/j.ipm.2023.103564 · Cited by 140
Correcting the Triplet Selection Bias for Triplet Loss
Yu, B., Liu, T., Gong, M., Ding, C., & Tao, D. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01231-1_5 · Cited by 127
Enhancing remote sensing image retrieval using a triplet deep metric learning network
Cao, R., Zhang, Q., Zhu, J., Li, Q., Li, Q., Liu, B., & Qiu, G. (2019) · International Journal of Remote Sensing · doi:10.1080/2150704x.2019.1647368 · Cited by 112
“Deep metric learning is proven to be effective in fields like face recognition [6], person re-identification [7], and natural image retrieval [8].”
BackgroundJin, S., Su, Y., Gao, S., Wu, F., Ma, Q., Xu, K., Hu, T., Liu, J., Pang, S., Guan, H., et al. (2019) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2019.2953092 · Cited by 113
Khan, A., Ilyas, T., Umraiz, M., Mannan, Z. I., & Kim, H. (2020) · Electronics · doi:10.3390/electronics9101602 · Cited by 124
“This capability has enabled the expansion of CNN's applications to all fields of computer vision, including self-driving cars [13], facial recognition [14], stereo vision [15], medical image processing [16], agriculture [7], and bioinformatics [17].”
BackgroundMasked Face Recognition with Latent Part Detection
Ding, F., Peng, P., Huang, Y., Geng, M., & Tian, Y. (2020) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3394171.3413731 · Cited by 116
Clapham, M., Miller, E., Nguyen, M., & Darimont, C. T. (2020) · Ecology and Evolution · doi:10.1002/ece3.6840 · Cited by 115
“This stage uses the facial landmarks in the XML created by bear-face to reorientate and extract the bear faces (or “chips”: Schroff et al., 2015).”
MethodInfluentialGabor filter bank with deep autoencoder based face recognition system
Hammouche, R., Attıa, A., Akhrouf, S., & Akhtar, Z. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.116743 · Cited by 120
Enhancement of neural networks with an alternative activation function tanhLU
Shen, S., Zhang, N., Zhou, A., & Yin, Z. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.117181 · Cited by 124
Lv, J., Kuang, J., Yu, Z., Sun, G., Liu, J., & Leon, J. I. (2023) · IEEE Transactions on Industrial Electronics · doi:10.1109/tie.2023.3241404 · Cited by 110
“tuning network by triplet loss [22], which is always utilized to improve the robustness of the features in classification tasks.”
MethodSynthetic Examples Improve Generalization for Rare Classes
Beery, S., Liu, Y., Morris, D., Piavis, J., Kapoor, A., Meister, M., Joshi, N., & Perona, P. (2020) · 2020 IEEE Winter Conference on Applications of Computer Vision (WACV) · doi:10.1109/wacv45572.2020.9093570 · Cited by 103
“It remains to be studied whether embedding techniques [68], domain adaptation techniques [34, 84], or style transfer [35, 70] could be used to encourage a higher overlap in representation between the synthetic and real data, and whether that overlap would in fact lead to an increase in…”
BackgroundSaliency Detection in 360$$^\circ $$ Videos
Zhang, Z., Xu, Y., Jingyi Yu, A., et al. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01234-2_30 · Cited by 105
Daedalus: Breaking Nonmaximum Suppression in Object Detection via Adversarial Examples
Wang, D., Li, C., Wen, S., Han, Q., Nepal, ., Zhang, X., & Xiang, Y. (2021) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2020.3041481 · Cited by 144
Chen, Y., Gan, H., Chen, H., Zeng, Y., Xu, L., Heidari, A. A., Zhu, X., & Liu, Y. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.10.064 · Cited by 112
Toward More Accurate Iris Recognition Using Dilated Residual Features
Wang, K., & Kumar, A. (2019) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2019.2913234 · Cited by 110
“the training process with offline triplets selection [24].”
MethodLearning SO(3) Equivariant Representations with Spherical CNNs
Esteves, C., Allen-Blanchette, C., Makadia, A., & Daniilidis, K. (2019) · International Journal of Computer Vision · doi:10.1007/s11263-019-01220-1 · Cited by 157
“The network is trained for classification on the 55 core classes (we do not use the subclasses), with an extra in-batch triplet loss (from [40]) to encourage descriptors to be close for matching categories and far for non-matching.”
MethodMulti-Modal Multi-Channel Target Speech Separation
Gu, R., Zhang, S.-X., Xu, Y., Chen, L., Zou, Y., & Yu, D. (2020) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2020.2980956 · Cited by 146
“Moreover, the face indicates the speaker identity information [15].”
BackgroundClusterFit: Improving Generalization of Visual Representations
Yan, X., Misra, I., Gupta, A., Ghadiyaram, D., & Mahajan, D. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00654 · Cited by 98
Perspectives on Individual Animal Identification from Biology and Computer Vision
Vidal, M., Wolf, N., Rosenberg, B., Harris, B. P., & Mathis, A. (2021) · Integrative and Comparative Biology · doi:10.1093/icb/icab107 · Cited by 104
“(58) which used triplet loss with semi hard mining on large batches.”
BackgroundGait Recognition in the Wild: A Benchmark
(2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01452 · Cited by 252
SFace: Privacy-friendly and Accurate Face Recognition using Synthetic Data
Boutros, F., Huber, M., Siebke, P., Rieber, T., & Damer, N. (2022) · IEEE International Joint Conference on Biometrics (IJCB) · doi:10.1109/ijcb54206.2022.10007961 · Cited by 164
“1% on LFW using FaceNet [29] trained on a synthetically generated dataset.”
MethodInfluentialVision-Based UAV Self-Positioning in Low-Altitude Urban Environments
Dai, M., Zheng, E., Feng, Z., Qi, L., Zhuang, J., & Yang, W. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3346279 · Cited by 168
“To evaluate the effectiveness of metric learning, we conduct experiments using four types of triplet loss, including the standard Triplet Loss [40], Hard-Mining Triplet Loss , Same-Domain Triplet Loss [3], and Soft-Weighted Triplet Loss [41].”
MethodFAREC — CNN based efficient face recognition technique using Dlib
Sharma, S., Shanmugasundaram, K., & Ramasamy, S. K. (2016) · IEEE International Conference on Advanced Communications, Control and Computing Technologies · doi:10.1109/icaccct.2016.7831628 · Cited by 121
Zeng, X., Cao, K., & Zhang, M. (2017) · Proceedings of the 15th Annual International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3081333.3081336 · Cited by 103
Multimodal and Temporal Perception of Audio-visual Cues for Emotion Recognition
Ghaleb, E., Popa, M., & Asteriadis, S. (2019) · 2019 8th International Conference on Affective Computing and Intelligent Interaction (ACII) · doi:10.1109/acii.2019.8925444 · Cited by 93
“the distance from the anchor to the negative pair [26].”
BackgroundDeep learning-based intelligent face recognition in IoT-cloud environment
Masud, M., Muhammad, G., Alhumyani, H. A., Alshamrani, S. S., Cheikhrouhou, O., Ibrahim, S., & Hossain, M. S. (2020) · Computer Communications · doi:10.1016/j.comcom.2020.01.050 · Cited by 107
“Google introduced the FaceNet [33], which used GoogleNet-24 architecture, got more than 99% accuracy on the same database.”
MethodResultInfluentialMasked Face Recognition Using Convolutional Neural Network
Ejaz, M. S., & Islam, M. R. (2019) · International Conference on Sustainable Technologies for Industry 5.0 (STI) · doi:10.1109/sti47673.2019.9068044 · Cited by 118
“Within those algorithms, the leading accuracy holds FaceNet and at the LFW dataset, it reached a 99.63% accuracy rate, which is higher than the normal human eye [9].”
BackgroundInfluentialBui, N. D. Q., Yu, Y., & Jiang, L. (2021) · Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval · doi:10.1145/3404835.3462840 · Cited by 103
“Contrastive learning has emerged as a new paradigm that brings together multiple forms of supervised learning problem as the task to compare similar and dissimilar items, such as siamese neural networks [9], contrastive predictive coding [57], triplet loss [61].”
BackgroundIntelligent diagnosis of northern corn leaf blight with deep learning model
Shuai-qun, P. A. N., Qiao, J., Wang, R., Yu, H., Wang, C., Taylor, K. L., & Pan, H. (2022) · Journal of Integrative Agriculture · doi:10.1016/s2095-3119(21)63707-3 · Cited by 117
Dynamic Prototype Mask for Occluded Person Re-Identification
Tan, L., Dai, P., Ji, R., & Wu, Y. (2022) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3503161.3547764 · Cited by 135
“To increase the intra-class similarity and decrease the interclass similarity, triplet loss Ltri with online hard-mining [30] are combined during the supervise training.”
MethodSelf-Supervised Remote Sensing Feature Learning: Learning Paradigms, Challenges, and Future Works
Tao, C., Qi, J., Guo, M., Zhu, Q., & Li, H. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3276853 · Cited by 125
“[147], [148], [149], [150], [151], [152], [153], [154] takes different augmented views of the same image as positive samples and the views of any other images as negative”
BackgroundMSINet: Twins Contrastive Search of Multi-Scale Interaction for Object ReID
Gu, J., Wang, K., Luo, H., Chen, C., Jiang, W., Fang, Y., Zhang, S., You, Y., & Zhao, J. G. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.01844 · Cited by 129
“Moreover, ReID is required to distinguish more subtle distinctions among fine-grained instances compared with image-level classification [48, 63].”
MethodStructure Invariant Transformation for better Adversarial Transferability
Wang, X., Zhang, Z., & Zhang, J. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00425 · Cited by 187
Generative Models for Effective ML on Private, Decentralized Datasets
Augenstein, S., McMahan, H. B., Ramage, D., Ramaswamy, S., Kairouz, P., Chen, M., Mathews, R., & Arcas, B. A. Y. (2019) · arXiv · doi:10.48550/arxiv.1911.06679 · Cited by 103
“An example is the triplet loss used for FaceNet in Schroff et al. (2015).”
MethodA Comprehensive Study on Center Loss for Deep Face Recognition
Wen, Y., Zhang, K., Li, Z., & Qiao, Y. (2019) · International Journal of Computer Vision · doi:10.1007/s11263-018-01142-4 · Cited by 113
“2016a) and triplet loss (Schroff et al. 2015) are proposed to enhance the discriminative power of the deeply learned face features.”
BackgroundInfluentialDeep Metric Learning With Tuplet Margin Loss
Yu, B., & Tao, D. (2019) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00659 · Cited by 130
“To solve this problem, we introduce a slack margin in angular space to pay more attention to “moderately hard triplets” rather than “the hardest triplets” [23].”
BackgroundMethodInfluentialJRMOT: A Real-Time 3D Multi-Object Tracker and a New Large-Scale Dataset
Shenoi, A., Patel, M., Gwak, J., Goebel, P., Sadeghian, A., Rezatofighi, H., Martín-Martín, R., & Savarese, S. (2020) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros45743.2020.9341635 · Cited by 124
“We train this fusion network via metric learning based on the triplet loss and semi-hard negative mining as in Schroff et al. [27], resulting in a feature useful for robust association between new detections and previous tracks.”
MethodForensics and Analysis of Deepfake Videos
Jafar, M. T., Ababneh, M., Al-Zoube, M., & Elhassan, A. (2020) · 2020 11th International Conference on Information and Communication Systems (ICICS) · doi:10.1109/icics49469.2020.239493 · Cited by 102
“Deepfake Video-TIMIT dataset was evaluated by implementing VGG [17] and Facenet [18] neural network algorithms in order to classify deepfake videos and also to detect discrepancies between lip movement and voice synchronization in both videos.”
MethodA Structured Model for Action Detection
Zhang, Y., Tokmakov, P., Hebert, M., & Schmid, C. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.01021 · Cited by 125
“To learn an appearance feature encoding for distinguishing different actors, we train a Siamese network [19] with a triplet loss [46].”
MethodFace attribute prediction using off-the-shelf CNN features
Zhong, Y., Sullivan, J., & Li, H. (2016) · International Conference on Biometrics · doi:10.1109/icb.2016.7550092 · Cited by 116
“Besides the DeepID series approaches [18, 16, 17], related efforts have also been made to pose correction [20], architectures design [19, 14] and data collection skills [12].”
MethodPinaya, W. H. L., Vieira, S., Garcia‐Dias, R., & Mechelli, A. (2019) · Elsevier eBooks · doi:10.1016/b978-0-12-815739-8.00010-9 · Cited by 172
Noise Estimation Using Density Estimation for Self-Supervised Multimodal Learning
Amrani, E., Ben‐Ari, R., Rotman, D., & Bronstein, A. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i8.16822 · Cited by 83
FashionVLP: Vision Language Transformer for Fashion Retrieval with Feedback
Goenka, S., Zheng, Z., Jaiswal, A., Chada, R., Wu, Y., Hedau, V., & Natarajan, P. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.01371 · Cited by 159
A Survey on Evaluation Metrics for Machine Translation
Lee, S., Lee, J., Moon, H., Park, C., Seo, J., Eo, S., Koo, S., & Lim, H. (2023) · Mathematics · doi:10.3390/math11041006 · Cited by 176
Quality-aware Pretrained Models for Blind Image Quality Assessment
Zhao, K., Yuan, K., Sun, M. B., Li, M., & Wen, X. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.02136 · Cited by 163
Rethinking Domain Generalization for Face Anti-spoofing: Separability and Alignment
Sun, Y., Liu, Y., Liu, X., Li, Y., & Chu, W. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.02353 · Cited by 156
Spatial–Temporal Attention Network for Depression Recognition from facial videos
Pan, Y., Shang, Y., Liu, T., Shao, Z., Guo, G., Ding, H., & Hu, Q. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.121410 · Cited by 121
Face Recognition: From Traditional to Deep Learning MethodsPreprint
Sáez-Trigueros, D., Meng, L., & Hartnett, M. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.00116 · Cited by 187
Learning Pose-Aware Models for Pose-Invariant Face Recognition in the Wild
Masi, I., Chang, F.-J., Choi, J., Harel, S., Kim, J., Kim, K., Leksut, J. T., Rawls, S., Wu, Y., Hassner, T., et al. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2018.2792452 · Cited by 115
Speech-Driven Facial Animation Using Cascaded GANs for Learning of Motion and Texture
Das, D., Biswas, S., Sinha, S., & Bhowmick, B. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58577-8_25 · Cited by 141
“We compute the performance metrics PSNR, SSIM (Structural Similarity), CPBD (Cumulative Probability Blur Detection) [25], ACD (Average Content Distance) [32] and similarity between FaceNet [27] features for reference identity image (1st frame of ground truth video) and the predicted frames.”
MethodA Survey on Adversarial Domain Adaptation
Zonoozi, M. H. P., & Seydi, V. (2022) · Neural Processing Letters · doi:10.1007/s11063-022-10977-5 · Cited by 133
“loss [60], the contrastive loss [64] and the center loss [81].”
BackgroundRice, G., Wagner, T., Stabrin, M., Sitsel, O., Prumbaum, D., & Raunser, S. (2023) · Nature Methods · doi:10.1038/s41592-023-01878-z · Cited by 118
Bluwstein, K., Buckmann, M., Joseph, A. C., Kapadia, S., & Şimşek, Ö. (2023) · Journal of International Economics · doi:10.1016/j.jinteco.2023.103773 · Cited by 152
Channel Augmentation for Visible-Infrared Re-Identification
Ye, M., Wu, Z., Chen, C., & Du, B. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3332875 · Cited by 132
“(6), (cid:2) 1 norm [86], [87], [88] is widely regarded as a technique for measuring pairwise distance difference.”
Self-supervised anomaly detection in computer vision and beyond: A survey and outlook
Hojjati, H., Ho, T. K. K., & Armanfard, N. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.106106 · Cited by 174
Deep Learning in Object Recognition, Detection, and Segmentation
Wang, X. (2016) · Foundations and Trends® in Signal Processing · doi:10.1561/2000000071 · Cited by 128
“Now the new state-of-the-art DeepID2+ [137] and FaceNet [123] have achieved face verification accuracy of 99.”
BackgroundHierarchy-Based Image Embeddings for Semantic Image Retrieval
Barz, B., & Denzler, J. (2019) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2019.00073 · Cited by 134
Metadata
- Authors
- Schroff, F., Kalenichenko, D., & Philbin, J.
- Affiliations
- Google (United States) (US)
- Venue
- 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2015-06
- Pages
- 815-823
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W2096733369
- Semantic Scholar
- 5aa26299435bdf7db874ef1640a6c3b5a4a2c394
- Cited by
- 15,267
- References
- 22
- Citations by year
- 2026: 846 · 2025: 1556 · 2024: 1590 · 2023: 1823 · 2022: 1872 · 2021: 2113 · 2020: 1902 · 2019: 1574 · 2018: 1062 · 2017: 593 · 2016: 281 · 2015: 51
- FWCI
- 271.65
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- arXiv (Cornell University) (publishedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face recognition and analysis
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face and Expression Recognition
- Life Sciences › Neuroscience › Cognitive Neuroscience › Face Recognition and Perception
- Keywords
- FaceNet, face recognition, face verification, face clustering, embedding space, triplet loss, deep convolutional neural networks, face similarity, Labeled Faces in the Wild
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.