3D face modeling has been an active area of research in computer vision and computer graphics, fueling applications ranging from facial expression transfer in virtual avatars to synthetic data generation. Existing 3D deep learning generative models (e.g., VAE, GANs) allow generating compact face representations (both shape and texture) that can model non-linearities in the shape and appearance space (e.g., scatter effects, specularities,..). However, they lack the capability to control the generation of subtle expressions. This paper proposes a new 3D face generative model that can decouple identity and expression and provides granular control over expressions. In particular, we propose using a pair of supervised auto-encoder and generative adversarial networks to produce high-quality 3D faces, both in terms of appearance and shape. Experimental results in the generation of 3D faces learned with holistic expression labels, or Action Unit (AU) labels, show how we can decouple identity and expression; gaining fine-control over expressions while preserving identity.1

This paper presents a supervised mixing augmentation method termed SuperMix, which exploits the salient regions within input images to construct mixed training samples. SuperMix is designed to obtain mixed images rich in visual features and complying with realistic image priors. To enhance the efficiency of the algorithm, we develop a variant of the Newton iterative method, 65× faster than gradient descent on this problem. We validate the effectiveness of SuperMix through extensive evaluations and ablation studies on two tasks of object classification and knowledge distillation. On the classification task, SuperMix provides comparable performance to the advanced augmentation methods, such as AutoAugment and RandAugment. In particular, combining SuperMix with RandAugment achieves 78.2% top-1 accuracy on ImageNet with ResNet50. On the distillation task, solely classifying images mixed using the teacher’s knowledge achieves comparable performance to the state-of-the-art distillation methods. Furthermore, on average, incorporating mixed images into the distillation objective improves the performance by 3.4% and 3.1% on CIFAR-100 and ImageNet, respectively. The code is available at https://github.com/alldbi/SuperMix.

The goal is to use Wasserstein metric to provide pseudo labels for the unlabeled images to train a Convolutional Neural Networks (CNN) in a Semi-Supervised Learning (SSL) manner for the classification task. The basic premise in our method is that the discrepancy between two discrete empirical measures (e.g., clusters) which come from the same or similar distribution is expected to be less than the case where these measures come from completely two different distributions. In our proposed method, we first pre-train our CNN using a self-supervised learning method to make a cluster assumption on the unlabeled images. Next, inspired by the Wasserstein metric which considers the geometry of the metric space to provide a natural notion of similarity between discrete empirical measures, we leverage it to cluster the unlabeled images and then match the clusters to their similar class of labeled images to provide a pseudo label for the data within each cluster. We have evaluated and compared our method with state-of-the-art SSL methods on the standard datasets to demonstrate its effectiveness.

In this paper, we present a novel differential morph detection framework, utilizing landmark and appearance disentanglement. In our framework, the face image is represented in the embedding domain using two disentangled but complementary representations. The network is trained by triplets of face images, in which the intermediate image inherits the landmarks from one image and the appearance from the other image. This initially trained network is further trained for each dataset using contrastive representations. We demonstrate that, by employing appearance and landmark disentanglement, the proposed frame-work can provide state-of-the-art differential morph detection performance. This functionality is achieved by the using distances in landmark, appearance, and ID domains. The performance of the proposed framework is evaluated using three morph datasets generated with different methodologies.

Recently, ensemble models have demonstrated empirical capabilities to alleviate the adversarial vulnerability. In this paper, we exploit first-order interactions within ensembles to formalize a reliable and practical defense. We introduce a scenario of interactions that certifiably improves the robustness according to the size of the ensemble, the diversity of the gradient directions, and the balance of the member's contribution to the robustness. We present a joint gradient phase and magnitude regularization (GPMR) as a vigorous approach to impose the desired scenario of interactions among members of the ensemble. Through extensive experiments, including gradient-based and gradient-free evaluations on several datasets and network architectures, we validate the practical effectiveness of the proposed approach compared to the previous methods. Furthermore, we demonstrate that GPMR is orthogonal to other defense strategies developed for single classifiers and their combination can further improve the robustness of ensembles.

Cross-spectral iris recognition is emerging as a promising biometric approach to authenticating the identity of individuals. However, matching iris images acquired at different spectral bands shows significant performance degradation when compared to single-band near-infrared (NIR) matching due to the spectral gap between iris images obtained in the NIR and visual-light (VIS) spectra. Although researchers have recently focused on deep-learning-based approaches to recover invariant representative features for more accurate recognition performance, the existing methods cannot achieve the expected accuracy required for commercial applications. Hence, in this paper, we propose a conditional coupled generative adversarial network (CpGAN) architecture for cross-spectral iris recognition by projecting the VIS and NIR iris images into a low-dimensional embedding domain to explore the hidden relationship between them. The conditional CpGAN framework consists of a pair of GAN-based networks, one responsible for retrieving images in the visible domain and other responsible for retrieving images in the NIR domain. Both networks try to map the data into a common embedding subspace to ensure maximum pair-wise similarity between the feature vectors from the two iris modalities of the same subject. To prove the usefulness of our proposed approach, extensive experimental results obtained on the PolyU dataset are compared to existing state-of-the-art cross-spectral recognition methods.

In recent years, due to the emergence of deep learning, face recognition has achieved exceptional success. However, many of these deep face recognition models perform relatively poorly in handling profile faces compared to frontal faces. The major reason for this poor performance is that it is inherently difficult to learn large pose invariant deep representations that are useful for profile face recognition. In this paper, we hypothesize that the profile face domain possesses a gradual connection with the frontal face domain in the deep feature space. We look to exploit this connection by projecting the profile faces and frontal faces into a common latent space and perform verification or retrieval in the latent domain. We leverage a coupled generative adversarial network (cpGAN) structure to find the hidden relationship between the profile and frontal images in a latent common embedding subspace. Specifically, the cp-GAN framework consists of two GAN-based sub-networks, one dedicated to the frontal domain and the other dedicated to the profile domain. Each sub-network tends to find a projection that maximizes the pair-wise correlation between two feature domains in a common embedding feature subspace. The efficacy of our approach compared with the state-of-the-art is demonstrated using the CFp, CMU Multi-PIE, IJB-A, and IJB-C datasets.

Deep neural networks are susceptible to adversarial manipulations in the input domain. The extent of vulnerability has been explored intensively in cases of ℓp-bounded and ℓp-minimal adversarial perturbations. However, the vulnerability of DNNs to adversarial perturbations with specific statistical properties or frequency-domain characteristics has not been sufficiently explored. In this paper, we study the smoothness of perturbations and propose Smooth-Fool, a general and computationally efficient framework for computing smooth adversarial perturbations. Through extensive experiments, we validate the efficacy of the proposed method for both the white-box and black-box attack scenarios. In particular, we demonstrate that: (i) there exist extremely smooth adversarial perturbations for well-established and widely used network architectures, (ii) smoothness significantly enhances the robustness of perturbations against state-of-the-art defense mechanisms, (iii) smoothness improves the transferability of adversarial perturbations across both data points and network architectures, and (iv) class categories exhibit a variable range of susceptibility to smooth perturbations. Our results suggest that smooth APs can play a significant role in exploring the vulnerability extent of DNNs to adversarial examples. The code is available at https://github.com/alldbi/SmoothFool.

In this paper, we propose a framework for disentangling the appearance and geometry representations in the face recognition task. To provide supervision for this aim, we generate geometrically identical faces by incorporating spatial transformations. We demonstrate that the proposed approach enhances the performance of deep face recognition models by assisting the training process in two ways. First, it enforces the early and intermediate convolutional layers to learn more representative features that satisfy the properties of disentangled embeddings. Second, it augments the training set by altering faces geometrically. Through extensive experiments, we demonstrate that integrating the proposed approach into state-of-the-art face recognition methods effectively improves their performance on challenging datasets, such as LFW, YTF, and MegaFace. Both theoretical and practical aspects of the method are analyzed rigorously by concerning ablation studies and knowledge transfer tasks. Furthermore, we show that the knowledge leaned by the proposed method can favor other face-related tasks, such as attribute prediction.

Deep generative models are a set of promising methods, that are able to model complex data and generate new samples. In principle, they learn to map a random latent code sampled from a prior distribution into a high dimensional data space, such as image space. However, these models have limited utilities as the user has minimal control over what the network produces. Despite the success of some recent work in learning an interpretable latent code, the field still lacks a coherent framework to learn a fully interpretable latent code, without any random part for sample diversity. Consequently, it is generally hard, if not impossible, for a non-expert user to produce a desired image by tuning the random and interpretable parts of the latent code. In this paper, we introduce the Preference-Based Image Generation (PbIG), a new method to retrieve the corresponding latent code of the user's mental image. We propose to adopt preference-based reinforcement learning, which learns from a user's judgment of the generated images by a pre-trained generative model. Since the proposed method is completely decoupled from the training stage of the underlying generative models, it can easily be adopted by any method, such as GANs and VAEs. We evaluate the effectiveness of PbIG framework using a set of experiments on baseline datasets using a pretraind StackGAN++.

Convolutional Neural Networks (CNNs) have provided promising achievements for image classification problems. However, training a CNN model relies on a large number of labeled data. Considering the vast amount of unlabeled data available on the web, it is important to make use of these data in conjunction with a small set of labeled data to train a deep learning model. In this paper, we introduce a new iterative Graph-based Semi-Supervised Learning (GSSL) method to train a CNN-based classifier using a large amount of unlabeled data and a small amount of labeled data. In this method, we first construct a similarity graph in which the nodes represent the CNN features corresponding to data points (labeled and unlabeled) while the edges tend to connect the data points with the same class label. In this graph, the missing label of unsupervised nodes is predicted by using a matrix completion method based on rank minimization criterion. In the next step, we use the constructed graph to calculate triplet regularization loss which is added to the supervised loss obtained by initially labeled data to update the CNN network parameters.

Objects are usually organized in a hierarchical structure in which each coarse category (e.g., big cat) corresponds to a superclass of several fine categories (e.g., cheetah, leopard). The objects grouped within the same coarse category, but in different fine categories, usually share a set of global visual features; however, these objects have distinctive local properties that characterize them at a fine level. This paper addresses the challenge of fine image classification in a weakly supervised fashion, whereby a subset of images is tagged by fine labels, while the remaining are tagged by coarse labels. We propose a new deep model that leverages coarse images to improve the classification performance of fine images within the coarse category. Our model is an end to end framework consisting of a Convolutional Neural Network (CNN) which uses both fine and coarse images to tune its parameters. The CNN outputs are then fanned out into two separate branches such that the first branch uses a supervised low rank self expressive layer to project the CNN outputs to the low rank subspaces to capture the global structures for the coarse classification, while the other branch uses a supervised sparse self expressive layer to project them to the sparse subspaces to capture the local structures for the fine classification. Our deep model uses coarse images in conjunction with fine images to jointly explore the low rank and sparse subspaces by sharing the parameters during the training which causes the data points obtained by the CNN to be well-projected to both sparse and low rank subspaces for classification.

Unsupervised image-to-image translation is a class of computer vision problems which aims at modeling conditional distribution of images in the target domain, given a set of unpaired images in the source and target domains. An image in the source domain might have multiple representations in the target domain. Therefore, ambiguity in modeling of the conditional distribution arises, specially when the images in the source and target domains come from different modalities. Current approaches mostly rely on simplifying assumptions to map both domains into a shared-latent space. Consequently, they are only able to model the domain-invariant information between the two modalities. These approaches cannot model domain-specific information which has no representation in the target domain. In this work, we propose an unsupervised image-to-image translation framework which maximizes a domain-specific variational information bound and learns the target domain-invariant representation of the two domain. The proposed framework makes it possible to map a single source image into multiple images in the target domain, utilizing several target domain-specific codes sampled randomly from the prior distribution, or extracted from reference images.

Preserving details while restoring images highly corrupted by impulsive salt and pepper noise remains a challenging problem. The authors proposed an algorithm based on radial basis functions (RBFs) interpolation which estimates the intensities of corrupted pixels by their neighbours. In this algorithm, intensity values of noisy pixels in the corrupted image are first estimated using RBFs. Next, the image is smoothed. The proposed algorithm can effectively remove the highly dense, impulsive salt and pepper noise. Experimental results show the superiority of the proposed algorithm both in noise suppression and details preservation in comparison to the recent similar methods. Extensive simulations show better results measured by peak signal‐to‐noise ratio and structural similarity index, especially when the image is corrupted by very highly dense impulse noise.