# SphereFace2: Binary Classification is All You Need for Deep Face Recognition

arXiv: https://arxiv.org/abs/2108.01513 · HTML: https://mlsanity.com/paper/2108.01513/

## Key facts

- Cited by 42 papers in the map (data through 14 Oct 2026).
- Builds on 33 of its 52 references that are in the map.
- First posted on arXiv 3 Aug 2021.

## Builds on (most cited first, 30 of 33)

- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [FaceNet: A Unified Embedding for Face Recognition and Clustering](https://mlsanity.com/paper/1503.03832/) (2015)
- [CosFace: Large Margin Cosine Loss for Deep Face Recognition](https://mlsanity.com/paper/1801.09414/) (2018)
- [VGGFace2: A dataset for recognising faces across pose and age](https://mlsanity.com/paper/1710.08092/) (2017)
- [Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks](https://mlsanity.com/paper/1604.02878/) (2016)
- [SphereFace: Deep Hypersphere Embedding for Face Recognition](https://mlsanity.com/paper/1704.08063/) (2017)
- [Deep Metric Learning via Lifted Structured Feature Embedding](https://mlsanity.com/paper/1511.06452/) (2015)
- [Circle Loss: A Unified Perspective of Pair Similarity Optimization](https://mlsanity.com/paper/2002.10857/) (2020)
- [Multi-Similarity Loss with General Pair Weighting for Deep Metric Learning](https://mlsanity.com/paper/1904.06627/) (2019)
- [CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition](https://mlsanity.com/paper/2004.00288/) (2020)
- [Large-Margin Softmax Loss for Convolutional Neural Networks](https://mlsanity.com/paper/1612.02295/) (2016)
- [Sampling Matters in Deep Embedding Learning](https://mlsanity.com/paper/1706.07567/) (2017)
- [Additive Margin Softmax for Face Verification](https://mlsanity.com/paper/1801.05599/) (2018)
- [Cross-Age LFW: A Database for Studying Cross-Age Face Recognition in Unconstrained Environments](https://mlsanity.com/paper/1708.08197/) (2017)
- [NormFace: L2 Hypersphere Embedding for Face Verification](https://mlsanity.com/paper/1704.06369/) (2017)
- [The MegaFace Benchmark: 1 Million Faces for Recognition at Scale](https://mlsanity.com/paper/1512.00596/) (2015)
- [Partial FC: Training 10 Million Identities on a Single Machine](https://mlsanity.com/paper/2010.05222/) (2020)
- [Deep Learning Face Representation by Joint Identification-Verification](https://mlsanity.com/paper/1406.4773/) (2014)
- [Deep Metric Learning with Angular Loss](https://mlsanity.com/paper/1708.01682/) (2017)
- [L2-constrained Softmax Loss for Discriminative Face Verification](https://mlsanity.com/paper/1703.09507/) (2017)
- [Learning towards Minimum Hyperspherical Energy](https://mlsanity.com/paper/1805.09298/) (2018)
- [Deeply learned face representations are sparse, selective, and robust](https://mlsanity.com/paper/1412.1265/) (2014)
- [Deep Hyperspherical Learning](https://mlsanity.com/paper/1711.03189/) (2017)
- [Learning Deep Embeddings with Histogram Loss](https://mlsanity.com/paper/1611.00822/) (2016)
- [Orthogonal Over-Parameterized Training](https://mlsanity.com/paper/2004.04690/) (2020)
- [Ring loss: Convex Feature Normalization for Face Recognition](https://mlsanity.com/paper/1803.00130/) (2018)
- [Learning with Hyperspherical Uniformity](https://mlsanity.com/paper/2103.01649/) (2021)
- [Adversarial Learning with Margin-based Triplet Embedding Regularization](https://mlsanity.com/paper/1909.09481/) (2019)
- [Regularizing Neural Networks via Minimizing Hyperspherical Energy](https://mlsanity.com/paper/1906.04892/) (2019)
- [Multicolumn Networks for Face Recognition](https://mlsanity.com/paper/1807.09192/) (2018)

## Built on by (most cited first, 30 of 42)

- [Finetuning Text-to-Image Diffusion Models for Fairness](https://mlsanity.com/paper/2311.07604/) (2023)
- [EgoObjects: A Large-Scale Egocentric Dataset for Fine-Grained Object Understanding](https://mlsanity.com/paper/2309.08816/) (2023)
- [Benchmarking Algorithmic Bias in Face Recognition: An Experimental Approach Using Synthetic Faces and Human Evaluation](https://mlsanity.com/paper/2308.05441/) (2023)
- [SphereFace Revived: Unifying Hyperspherical Face Recognition](https://mlsanity.com/paper/2109.05565/) (2021)
- [Learning to Generate Image Embeddings with User-level Differential Privacy](https://mlsanity.com/paper/2211.10844/) (2022)
- [Part-based Face Recognition with Vision Transformers](https://mlsanity.com/paper/2212.00057/) (2022)
- [A Quality Aware Sample-to-Sample Comparison for Face Recognition](https://mlsanity.com/paper/2306.04000/) (2023)
- [FedFR: Joint Optimization Federated Framework for Generic and Personalized Face Recognition](https://mlsanity.com/paper/2112.12496/) (2021)
- [Generalizing and Decoupling Neural Collapse via Hyperspherical Uniformity Gap](https://mlsanity.com/paper/2303.06484/) (2023)
- [LAFS: Landmark-based Facial Self-supervised Learning for Face Recognition](https://mlsanity.com/paper/2403.08161/) (2024)
- [LVFace: Progressive Cluster Optimization for Large Vision Models in Face Recognition](https://mlsanity.com/paper/2501.13420/) (2025)
- [UniTSFace: Unified Threshold Integrated Sample-to-Sample Loss for Face Recognition](https://mlsanity.com/paper/2311.02523/) (2023)
- [CemiFace: Center-based Semi-hard Synthetic Face Generation for Face Recognition](https://mlsanity.com/paper/2409.18876/) (2024)
- [FedABC: Targeting Fair Competition in Personalized Federated Learning](https://mlsanity.com/paper/2302.07450/) (2023)
- [Cross-attention-based saliency inference for predicting cancer metastasis on whole slide images](https://mlsanity.com/paper/2309.09412/) (2023)
- [Combining Metric Learning and Attention Heads For Accurate and Efficient Multilabel Image Classification](https://mlsanity.com/paper/2209.06585/) (2022)
- [Pairwise Similarity Learning is SimPLE](https://mlsanity.com/paper/2310.09449/) (2023)
- [Learning Multi-Manifold Embedding for Out-Of-Distribution Detection](https://mlsanity.com/paper/2409.12479/) (2024)
- [IDFace: Face Template Protection for Efficient and Secure Identification](https://mlsanity.com/paper/2507.12050/) (2025)
- [Improving Bias Mitigation through Bias Experts in Natural Language Understanding](https://mlsanity.com/paper/2312.03577/) (2023)
- [Geometric Understanding of Discriminability and Transferability for Visual Domain Adaptation](https://mlsanity.com/paper/2407.09524/) (2024)
- [HyperSpaceX: Radial and Angular Exploration of HyperSpherical Dimensions](https://mlsanity.com/paper/2408.02494/) (2024)
- [GIF: Generative Inspiration for Face Recognition at Scale](https://mlsanity.com/paper/2505.03012/) (2025)
- [Non-Adaptive Adversarial Face Generation](https://mlsanity.com/paper/2507.12107/) (2025)
- [QCFace: Image Quality Control for boosting Face Representation & Recognition](https://mlsanity.com/paper/2510.15289/) (2025)
- [Spot keywords from very noisy and mixed speech](https://mlsanity.com/paper/2305.17706/) (2023)
- [Boosting Unconstrained Face Recognition with Targeted Style Adversary](https://mlsanity.com/paper/2408.07642/) (2024)
- [RepFace: Refining Closed-Set Noise with Progressive Label Correction for Face Recognition](https://mlsanity.com/paper/2412.12031/) (2024)
- [BCE vs. CE in Deep Feature Learning](https://mlsanity.com/paper/2505.05813/) (2025)
- [Adversarial Deep Metric Learning for Cross-Modal Audio-Text Alignment in Open-Vocabulary Keyword Spotting](https://mlsanity.com/paper/2505.16735/) (2025)

## Abstract (arXiv)

State-of-the-art deep face recognition methods are mostly trained with a softmax-based multi-class classification framework. Despite being popular and effective, these methods still have a few shortcomings that limit empirical performance. In this paper, we start by identifying the discrepancy between training and evaluation in the existing multi-class classification framework and then discuss the potential limitations caused by the "competitive" nature of softmax normalization. Motivated by these limitations, we propose a novel binary classification training framework, termed SphereFace2. In contrast to existing methods, SphereFace2 circumvents the softmax normalization, as well as the corresponding closed-set assumption. This effectively bridges the gap between training and evaluation, enabling the representations to be improved individually by each binary classification task. Besides designing a specific well-performing loss function, we summarize a few general principles for this "one-vs-all" binary classification framework so that it can outperform current competitive methods. Our experiments on popular benchmarks demonstrate that SphereFace2 can consistently outperform state-of-the-art deep face recognition methods. The code has been made publicly available.