CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition

@article{Huang2020CurricularFaceAC,
  title={CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition},
  author={Y. Huang and Yuhan Wang and Ying Tai and Xiaoming Liu and Pengcheng Shen and Shaoxin Li and Jilin Li and Feiyue Huang},
  journal={2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2020},
  pages={5900-5909}
}
  • Y. HuangYuhan Wang Feiyue Huang
  • Published 1 April 2020
  • Computer Science
  • 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
As an emerging topic in face recognition, designing margin-based loss functions can increase the feature margin between different classes for enhanced discriminability. More recently, the idea of mining-based strategies is adopted to emphasize the misclassified samples, achieving promising results. However, during the entire training process, the prior methods either do not explicitly emphasize the sample based on its importance that renders the hard samples not fully exploited; or explicitly… 
[PDF] Semantic Reader
249 Citations
Highly Influential Citations
45
Background Citations
116
Methods Citations
104
Results Citations
8

249 Citations

OTFace: Hard Samples Guided Optimal Transport Loss for Deep Face Representation

The hard samples guided optimal transport (OT) loss for deep face representation, OTFace, aims to enhance the performance of hard samples by introducing the feature distribution discrepancy while maintaining the performance on easy samples.

Variational Prototype Learning for Deep Face Recognition

This work identifies the slow feature drift phenomenon and directly inject memorized features into prototypes to approximate variational prototype sampling, which can simulate sample-to-sample comparisons within the classification framework, encouraging the SGD solver to be more exploratory, while boosting performance.

Sample hardness guided softmax loss for face recognition

This work develops a novel loss function, namely Hardness Loss, to adaptively assign weights for the misclassified (hard) samples guided by their corresponding hardness, which accounts for multiple training status and feature position information.

CoReFace: Sample-Guided Contrastive Regularization for Deep Face Recognition

This work proposes a novel approach namely Contrastive Regularization for Face recognition (CoReFace) to apply image-level regularization in feature representation learning and employs sample-guided contrastive learning to regularize the training with the image-image relationship directly, which is consistent with the evaluation process.

Sample-Level and Class-Level Adaptive Training for Face Recognition

This paper explicitly assign the adaptive margin according to the image quantity so that the margin is enlarged for the tail classes and semantically identify the ‘hard positive, samples and misclassified samples to attach adaptive weights to increase the training emphasis on these samples.

IHEM Loss: Intra-Class Hard Example Mining Loss for Robust Face Recognition

A hard example selection function is designed to effectively identify hard examples in the training procedure with the supervision of angular margin-based losses and a novel Intra-class Hard Example Mining (IHEM) loss function is proposed, which penalizes the cosine distance between the hard examples and their class centers to enhance the discriminative power of face representations.

SubFace: Learning with Softmax Approximation for Face Recognition

This paper presents a softmax approximation method called SubFace, which employs the subspace feature to promote the performance of face recognition, and dynamically selects the non-overlapping subspace features in each batch during training, and uses them to approximate full-feature among softmax-based loss.

KappaFace: Adaptive Additive Angular Margin Loss for Deep Face Recognition

This work introduces a novel adaptive strategy, called KappaFace, to modulate the relative importance based on class difficultness and imbalance, which can intensify the margin's magnitude for hard learning or low concentration classes while relaxing it for counter classes.

Curricular SincNet: Towards Robust Deep Speaker Recognition by Emphasizing Hard Samples in Latent Space

An improved SincNet model where a curricular loss function is used to train the Sinc net architecture, which achieves the best overall results with a reduction of 4% error rate compare to Sinc Net and other published work.

Improving Face Recognition from Hard Samples via Distribution Distillation Loss

This work proposes a novel Distribution Distillation Loss to narrow the performance gap between easy and hard samples, which is a simple, effective and generic for various types of facial variations.
...

45 References

AdaCos: Adaptively Scaling Cosine Logits for Effectively Learning Deep Face Representations

A novel cosine-based softmax loss is proposed, AdaCos, which is hyperparameter-free and leverages an adaptive scale parameter to automatically strengthen the training supervisions during the training process and outperforms state-of-the-art softmax losses on all the three datasets.

Mis-classified Vector Guided Softmax Loss for Face Recognition

This paper develops a novel loss function, which adaptively emphasizes the mis-classified feature vectors to guide the discriminative feature learning and is the first attempt to inherit the advantages of feature margin and feature mining into a unified loss function.

AdaptiveFace: Adaptive Margin and Sampling for Face Recognition

This paper proposes the Adaptive Margin Softmax to adjust the margins for different classes adaptively, and makes the sampling process adaptive in two folds: Firstly, the Hard Prototype Mining to adaptively select a small number of hard classes to participate in classification, and secondly, theAdaptive Data Sampling to find valuable samples for training adaptively.

Improving Face Recognition from Hard Samples via Distribution Distillation Loss

This work proposes a novel Distribution Distillation Loss to narrow the performance gap between easy and hard samples, which is a simple, effective and generic for various types of facial variations.

P2SGrad: Refined Gradients for Optimizing Deep Face Models

A novel gradient called P2SGrad (Probability-to-Similarity Gradient), which leverages a cosine similarity instead of classification probability to control the gradients for updating neural network parameters, which makes training process more efficient and faster.

Distribution Distillation Loss: Generic Approach for Improving Face Recognition from Hard Samples

This work proposes a novel Distribution Distillation Loss to narrow the performance gap between easy and hard samples, which is a simple, effective and generic for various types of facial variations.

A Discriminative Feature Learning Approach for Deep Face Recognition

This paper proposes a new supervision signal, called center loss, for face recognition task, which simultaneously learns a center for deep features of each class and penalizes the distances between the deep features and their corresponding class centers.

CosFace: Large Margin Cosine Loss for Deep Face Recognition

  • H. WangYitong Wang Wei Liu
  • Computer Science
    2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
  • 2018
This paper reformulates the softmax loss as a cosine loss by L2 normalizing both features and weight vectors to remove radial variations, based on which acosine margin term is introduced to further maximize the decision margin in the angular space, and achieves minimum intra-class variance and maximum inter- class variance by virtue of normalization and cosine decision margin maximization.

SphereFace: Deep Hypersphere Embedding for Face Recognition

This paper proposes the angular softmax (A-Softmax) loss that enables convolutional neural networks (CNNs) to learn angularly discriminative features in deep face recognition (FR) problem under open-set protocol.

ArcFace: Additive Angular Margin Loss for Deep Face Recognition

This paper presents arguably the most extensive experimental evaluation against all recent state-of-the-art face recognition methods on ten face recognition benchmarks, and shows that ArcFace consistently outperforms the state of the art and can be easily implemented with negligible computational overhead.