scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

ArcFace: Additive Angular Margin Loss for Deep Face Recognition

15 Jun 2019-pp 4690-4699
TL;DR: 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.
Abstract: One of the main challenges in feature learning using Deep Convolutional Neural Networks (DCNNs) for large-scale face recognition is the design of appropriate loss functions that can enhance the discriminative power. Centre loss penalises the distance between deep features and their corresponding class centres in the Euclidean space to achieve intra-class compactness. SphereFace assumes that the linear transformation matrix in the last fully connected layer can be used as a representation of the class centres in the angular space and therefore penalises the angles between deep features and their corresponding weights in a multiplicative way. Recently, a popular line of research is to incorporate margins in well-established loss functions in order to maximise face class separability. In this paper, we propose an Additive Angular Margin Loss (ArcFace) to obtain highly discriminative features for face recognition. The proposed ArcFace has a clear geometric interpretation due to its exact correspondence to geodesic distance on a hypersphere. We present arguably the most extensive experimental evaluation against all recent state-of-the-art face recognition methods on ten face recognition benchmarks which includes a new large-scale image database with trillions of pairs and a large-scale video dataset. We show that ArcFace consistently outperforms the state of the art and can be easily implemented with negligible computational overhead. To facilitate future research, the code has been made available.

Content maybe subject to copyright    Report

Citations
More filters
Proceedings ArticleDOI
27 Sep 2021
TL;DR: In this article, the authors investigated the effects of face subregions (semantic contexts) and link the general image quality of face regions with face image utility, and the effect of fusion of different face regions was investigated to increase the robustness of the outcomes.
Abstract: Face image quality assessment predicts the utility of a face image for automated face recognition. A high-quality face image can achieve good performance for the identification or verification task. Some recent face image quality assessment algorithms are established on deep-learning-based approaches, which rely on face embeddings of aligned face images. Such face embeddings fuse complex information into a single feature vector and are, therefore, challenging to disentangle. The semantic context however can provide better interpretable insights into neural-network decisions. We investigate the effects of face subregions (semantic contexts) and link the general image quality of face subregions with face image utility. The evaluation is performed on two difficult large-scale datasets (LFW and VGGFace2) with three face recognition solutions (FaceNet, SphereFace, and ArcFace). In total, we applied four face image quality assessment methods and one general image quality assessment method on four face subregions (eyes, mouth, nose, and tightly cropped face region) and the aligned faces. In addition, the effect of fusion of different face subregions was investigated to increase the robustness of the outcomes.

6 citations

Journal ArticleDOI
TL;DR: In this paper, a random interval attendance management system (RIAMS) is proposed for virtual learning in post-COVID settings, which employs a face recognition module built using the Dlib open-source software library.
Abstract: The exceptional circumstance caused by the COVID-19 pandemic demands substantial modifications in the teaching-learning processes across the globe. Teachers and students are making use of online learning in virtual classrooms as an alternative for face-to-face learning in physical classrooms. However, students’ attendance management during virtual learning is a challenging problem. It is quite difficult to identify students’ disengagement and even to know whether they are in front of their smart devices or not. In this paper, we introduce the ‘Random Interval Attendance Management System’ (RIAMS), which is an innovative solution for attendance monitoring issues, students’ disengagement, and attendance faking during virtual learning. In RIAMS, we employed a face recognition module built using the Dlib open-source software library. In order to improve the efficiency of the system, we introduced two ancillary modalities – verifying students’ responses to CAPTCHAs and UIN (Unique Identification Number) queries. Both the face recognition and ancillary modalities operate at random intervals of time. This distinctive feature of randomness in our design ensures that students’ attention and engagement in virtual learning are enhanced. Furthermore, the RIAMS’ multimodal architecture and its sub-modalities’ adaptive weight system enable teachers to customize their attendance strategy for every course. The output analysis of each of the RIAMS modalities and the combined results emphasize the effectiveness and reliability of our system in the attendance management for virtual learning. The novel RIAMS model has the potential to be extensively deployed for virtual learning in post-COVID settings.

6 citations

Journal ArticleDOI
TL;DR: In this paper , an implicit memory and an explicit memory are employed in the audio-to-expression model to capture high-level semantics in audio-expression shared space, while the explicit memory is employed in neural-rendering model to help synthesize pixel-level details.
Abstract: Talking face generation aims at generating photo-realistic video portraits of a target person driven by input audio. Due to its nature of one-to-many mapping from the input audio to the output video (e.g., one speech content may have multiple feasible visual appearances), learning a deterministic mapping like previous works brings ambiguity during training, and thus causes inferior visual results. Although this one-to-many mapping could be alleviated in part by a two-stage framework (i.e., an audio-to-expression model followed by a neural-rendering model), it is still insufficient since the prediction is produced without enough information (e.g., emotions, wrinkles, etc.). In this paper, we propose MemFace to complement the missing information with an implicit memory and an explicit memory that follow the sense of the two stages respectively. More specifically, the implicit memory is employed in the audio-to-expression model to capture high-level semantics in the audio-expression shared space, while the explicit memory is employed in the neural-rendering model to help synthesize pixel-level details. Our experimental results show that our proposed MemFace surpasses all the state-of-the-art results across multiple scenarios consistently and significantly.

6 citations

Journal ArticleDOI
TL;DR: In this paper, a multi-task collaboration network (MTCNet) is proposed for rotation-invariance face detection that fully uses facial landmarks to improve the detection performance by means of collaboration between face detection and face alignment.

6 citations

Posted Content
Ke Mei, Lei li, Xu Jinchang, Cheng Yanhua, Yugeng Lin 
TL;DR: This paper presents the 3rd place detailed solution to the Google Landmark Retrieval 2020 challenge using a data cleaning strategy based on embedding clustering and employing a data augmentation method called Corner-Cutmix, which improves the model's ability to recognize multi-scale and occluded landmark images.
Abstract: Image retrieval is a fundamental problem in computer vision. This paper presents our 3rd place detailed solution to the Google Landmark Retrieval 2020 challenge. We focus on the exploration of data cleaning and models with metric learning. We use a data cleaning strategy based on embedding clustering. Besides, we employ a data augmentation method called Corner-Cutmix, which improves the model's ability to recognize multi-scale and occluded landmark images. We show in detail the ablation experiments and results of our method.

6 citations


Cites methods from "ArcFace: Additive Angular Margin Lo..."

  • ...In this work, we first use ordinary softmax loss to pretrain a model on GLDv2 clean, and then use ArcFace [4] with a margin of 0....

    [...]

  • ...As shown in Table 1, we demonstrate the effectiveness of ArcFace, GLDv2 cluster and Corner-Cutmix....

    [...]

  • ...In this work, we first use ordinary softmax loss to pretrain a model on GLDv2 clean, and then use ArcFace [4] with a margin of 0.3 and scale of 30 to finetune it....

    [...]

References
More filters
Proceedings ArticleDOI
27 Jun 2016
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

123,388 citations

Journal Article
TL;DR: It is shown that dropout improves the performance of neural networks on supervised learning tasks in vision, speech recognition, document classification and computational biology, obtaining state-of-the-art results on many benchmark data sets.
Abstract: Deep neural nets with a large number of parameters are very powerful machine learning systems. However, overfitting is a serious problem in such networks. Large networks are also slow to use, making it difficult to deal with overfitting by combining the predictions of many different large neural nets at test time. Dropout is a technique for addressing this problem. The key idea is to randomly drop units (along with their connections) from the neural network during training. This prevents units from co-adapting too much. During training, dropout samples from an exponential number of different "thinned" networks. At test time, it is easy to approximate the effect of averaging the predictions of all these thinned networks by simply using a single unthinned network that has smaller weights. This significantly reduces overfitting and gives major improvements over other regularization methods. We show that dropout improves the performance of neural networks on supervised learning tasks in vision, speech recognition, document classification and computational biology, obtaining state-of-the-art results on many benchmark data sets.

33,597 citations

Proceedings Article
Sergey Ioffe1, Christian Szegedy1
06 Jul 2015
TL;DR: Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin.
Abstract: Training Deep Neural Networks is complicated by the fact that the distribution of each layer's inputs changes during training, as the parameters of the previous layers change. This slows down the training by requiring lower learning rates and careful parameter initialization, and makes it notoriously hard to train models with saturating nonlinearities. We refer to this phenomenon as internal covariate shift, and address the problem by normalizing layer inputs. Our method draws its strength from making normalization a part of the model architecture and performing the normalization for each training mini-batch. Batch Normalization allows us to use much higher learning rates and be less careful about initialization, and in some cases eliminates the need for Dropout. Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin. Using an ensemble of batch-normalized networks, we improve upon the best published result on ImageNet classification: reaching 4.82% top-5 test error, exceeding the accuracy of human raters.

30,843 citations

28 Oct 2017
TL;DR: An automatic differentiation module of PyTorch is described — a library designed to enable rapid research on machine learning models that focuses on differentiation of purely imperative programs, with a focus on extensibility and low overhead.
Abstract: In this article, we describe an automatic differentiation module of PyTorch — a library designed to enable rapid research on machine learning models. It builds upon a few projects, most notably Lua Torch, Chainer, and HIPS Autograd [4], and provides a high performance environment with easy access to automatic differentiation of models executed on different devices (CPU and GPU). To make prototyping easier, PyTorch does not follow the symbolic approach used in many other deep learning frameworks, but focuses on differentiation of purely imperative programs, with a focus on extensibility and low overhead. Note that this preprint is a draft of certain sections from an upcoming paper covering all PyTorch features.

13,268 citations

Posted Content
TL;DR: The TensorFlow interface and an implementation of that interface that is built at Google are described, which has been used for conducting research and for deploying machine learning systems into production across more than a dozen areas of computer science and other fields.
Abstract: TensorFlow is an interface for expressing machine learning algorithms, and an implementation for executing such algorithms. A computation expressed using TensorFlow can be executed with little or no change on a wide variety of heterogeneous systems, ranging from mobile devices such as phones and tablets up to large-scale distributed systems of hundreds of machines and thousands of computational devices such as GPU cards. The system is flexible and can be used to express a wide variety of algorithms, including training and inference algorithms for deep neural network models, and it has been used for conducting research and for deploying machine learning systems into production across more than a dozen areas of computer science and other fields, including speech recognition, computer vision, robotics, information retrieval, natural language processing, geographic information extraction, and computational drug discovery. This paper describes the TensorFlow interface and an implementation of that interface that we have built at Google. The TensorFlow API and a reference implementation were released as an open-source package under the Apache 2.0 license in November, 2015 and are available at www.tensorflow.org.

10,447 citations