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.
Citations
More filters
TL;DR: This work proposes three module variants using the SKA mechanism whereby two modules are applied in front of an ECAPA-TDNN model, and the other is combined with the Res2Net backbone block to outperforms the conventional counterpart on the three different evaluation protocols.
Abstract: Recent state-of-the-art speaker verification architectures adopt multi-scale processing and frequency-channel attention tech-niques. However, their full potential may not have been ex-ploited because these techniques’ receptive fields are fixed where most convolutional layers operate with specified kernel sizes such as 1, 3 or 5. We aim to further improve this line of research by introducing a selective kernel attention (SKA) mechanism. The SKA mechanism allows each convolutional layer to adaptively select the kernel size in a data-driven fashion based on an attention mechanism that exploits both frequency and channel domain using the previous layer’s output. We propose three module variants using the SKA mechanism whereby two modules are applied in front of an ECAPA-TDNN model, and the other is combined with the Res2Net backbone block. Experimental results demonstrate that our proposed model consistently outperforms the conventional counterpart on the three different evaluation protocols in terms of both equal error rate and minimum detection cost function. In addition, we present a detailed analysis that helps understand how the SKA module works.
3 citations
23 Aug 2022
TL;DR: The results show that trends appearing in a single-attribute analysis disappear or reverse when multi-attribute groups are considered, and that performance disparities are also related to non-protected attributes.
Abstract: In this paper, we propose a novel explanatory framework aimed to provide a better understanding of how face recognition models perform as the underlying data characteristics (protected attributes: gender, ethnicity, age; nonprotected attributes: facial hair, makeup, accessories, face orientation and occlusion, image distortion, emotions) on which they are tested change. With our framework, we evaluate ten state-of-the-art face recognition models, comparing their fairness in terms of security and usability on two data sets, involving six groups based on gender and ethnicity. We then analyze the impact of image characteristics on models performance. Our results show that trends appearing in a single-attribute analysis disappear or reverse when multi-attribute groups are considered, and that performance disparities are also related to non-protected attributes. Source code: https://cutt.1y/2XwRLiA.
3 citations
07 Sep 2022
TL;DR: This paper uses a coupled-encoder network to project frontal/profile face images into a common latent embedding space and learns to enlarge the margin between the distribution of genuine and imposter faces, which results in high mutual information between different views of the same identity.
Abstract: In this paper, we seek to draw connections between the frontal and profile face images in an abstract embedding space. We exploit this connection using a coupled-encoder network to project frontal/profile face images into a common latent embedding space. The proposed model forces the similarity of representations in the embedding space by maximizing the mutual information between two views of the face. The proposed coupled-encoder benefits from three contributions for matching faces with extreme pose disparities. First, we leverage our pose-aware contrastive learning to maximize the mutual information between frontal and profile representations of identities. Second, a memory buffer, which consists of latent representations accumulated over past iterations, is integrated into the model so it can refer to relatively much more instances than the mini-batch size. Third, a novel pose-aware adversarial domain adaptation method forces the model to learn an asymmetric mapping from profile to frontal representation. In our framework, the coupled-encoder learns to enlarge the margin between the distribution of genuine and imposter faces, which results in high mutual information between different views of the same identity. The effectiveness of the proposed model is investigated through extensive experiments, evaluations, and ablation studies on four benchmark datasets, and comparison with the compelling state-of-the-art algorithms.
3 citations
TL;DR: In this article , a text-conditioned fashion image editing model, called FICE (Fashion Image CLIP Editing), is proposed to handle a wide variety of semantically diverse text descriptions to guide the editing procedure.
Abstract: —Fashion–image editing represents a challenging com- puter vision task, where the goal is to incorporate selected apparel into a given input image. Most existing techniques deal with this task by first selecting an example image of the desired apparel and then transferring the clothing onto the target person. Such techniques are commonly referred to as Virtual Try-On methods. Conversely, in this paper, we consider editing fashion images with text descriptions. Such an approach has several advantages over example–based virtual try-on techniques, e.g.: ( i ) it does not require an image of the target fashion item, and ( ii ) it allows the expression of a wide variety of visual concepts through the use of natural language. Existing image–editing methods that work with language–based inputs are heavily constrained by their require- ment for training sets with rich attribute annotations or they are only able to handle simple text descriptions. To address these constraints, we propose a novel text–conditioned editing model, called FICE (Fashion Image CLIP Editing), capable of handling a wide variety of semantically diverse text descriptions to guide the editing procedure. Specifically with FICE, we augment the common GAN inversion process by including semantic, pose-related, and image-level constraints when generating the desired images. We leverage the capabilities of the pretrained CLIP model to enforce the targeted semantics, due to its impressive image–text association capabilities. We furthermore propose a latent–code regularization technique that provides the means to better control the fidelity of the synthesized images and ensures that images are generated from latent codes coming from a well-defined part of the latent space. We validate FICE through comprehensive experiments on a combination of VITON images and Fashion-Gen text descriptions and in comparison with several state-of-the-art text–conditioned image editing approaches. Experimental results demonstrate FICE generates highly realistic fashion images and leads to stronger editing performance than existing competing approaches.
3 citations
TL;DR: In this paper , a feed-forward convolutional encoder for the latent code and a fully-convolutional predictor of tri-plane numerical offsets are used for novel view rendering.
Abstract: Recent progress in NeRF-based GANs has introduced a number of approaches for high-resolution and high-fidelity generative modeling of human heads with a possibility for novel view rendering. At the same time, one must solve an inverse problem to be able to re-render or modify an existing image or video. Despite the success of universal optimization-based methods for 2D GAN inversion, those, applied to 3D GANs, may fail to produce 3D-consistent renderings. Fast encoder-based techniques, such as those developed for StyleGAN, may also be less appealing due to the lack of identity preservation. In our work, we introduce a real-time method that bridges the gap between the two approaches by directly utilizing the tri-plane representation introduced for EG3D generative model. In particular, we build upon a feed-forward convolutional encoder for the latent code and extend it with a fully-convolutional predictor of tri-plane numerical offsets. As shown in our work, the renderings are similar in quality to optimization-based techniques and significantly outperform the baselines for novel view. As we empirically prove, this is a consequence of directly operating in the tri-plane space, not in the GAN parameter space, while making use of an encoder-based trainable approach.
3 citations
References
More filters
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•
06 Jul 2015TL;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