Armon Matthew Safai
Bio: Armon Matthew Safai is an academic researcher from University of California, San Diego. The author has contributed to research in topics: Feature (machine learning) & Feature extraction. The author has an hindex of 3, co-authored 3 publications receiving 42 citations.
TL;DR: This work introduces three real-world open-set face recognition methods across facial plastic surgery changes and a look-alike face by 3D face reconstruction and sparse representation, and proposes likeness dictionary learning.
Abstract: The open-set problem is among the problems that have significantly changed the performance of face recognition algorithms in real-world scenarios. Open-set operates under the supposition that not all the probes have a pair in the gallery. Most face recognition systems in real-world scenarios focus on handling pose, expression and illumination problems on face recognition. In addition to these challenges, when the number of subjects is increased for face recognition, these problems are intensified by look-alike faces for which there are two subjects with lower intra-class variations. In such challenges, the inter-class similarity is higher than the intra-class variation for these two subjects. In fact, these look-alike faces can be created as intrinsic, situation-based and also by facial plastic surgery. This work introduces three real-world open-set face recognition methods across facial plastic surgery changes and a look-alike face by 3D face reconstruction and sparse representation. Since some real-world databases for face recognition do not have multiple images per person in the gallery, with just one image per subject in the gallery, this paper proposes a novel idea to overcome this challenge by 3D modeling from gallery images and synthesizing them for generating several images. Accordingly, a 3D model is initially reconstructed from frontal face images in a real-world gallery. Then, each 3D reconstructed face in the gallery is synthesized to several possible views and a sparse dictionary is generated based on the synthesized face image for each person. Also, a likeness dictionary is defined and its optimization problem is solved by the proposed method. Finally, the face recognition is performed for open-set face recognition using three proposed representation classifications. Promising results are achieved for face recognition across plastic surgery and look-alike faces on three databases including the plastic surgery face, look-alike face and LFW databases compared to several state-of-the-art methods. Also, several real-world and open-set scenarios are performed to evaluate the proposed method on these databases in real-world scenarios. This paper uses 3D reconstructed models to recognize look-alike faces.A feature is extracted from both facial reconstructed depth and texture images.This paper proposes likeness dictionary learning.Three open-set classification methods are proposed for real-world face recognition.
TL;DR: A novel method is proposed for Facial Expression Recognition (FER) using dictionary learning to learn both identity and expression dictionaries simultaneously and demonstrates excellent performance by obtaining high accuracy on all four databases but also outperforms other state-of-the-art approaches.
Abstract: Comprehensive feature extraction method is proposed for facial expression recognition.A sparse dictionary learning approach is proposed for facial expression recognition.A regression dictionary is proposed for regression facial expression classification.It improves the facial expression recognition rate on the CK+, MMI and JAFFE databases. In this paper, a novel method is proposed for Facial Expression Recognition (FER) using dictionary learning to learn both identity and expression dictionaries simultaneously. Accordingly, an automatic and comprehensive feature extraction method is proposed. The proposed method accommodates real-valued scores to a probability of what percent of the given Facial Expression (FE) is present in the input image. To this end, a dual dictionary learning method is proposed to learn both regression and feature dictionaries for FER. Then, two regression classification methods are proposed using a regression model formulated based on dictionary learning and two known classification methods including Sparse Representation Classification (SRC) and Collaborative Representation Classification (CRC). Convincing results are acquired for FER on the CK+, CK, MMI and JAFFE image databases compared to several state-of-the-arts. Also, promising results are obtained from evaluating the proposed method for generalization on other databases. The proposed method not only demonstrates excellent performance by obtaining high accuracy on all four databases but also outperforms other state-of-the-art approaches.
TL;DR: A regression model formulation is proposed for FAC in a wide range of FAs, which accommodates real-valued scores to the probability of what percentage of the given FAs is present in the input image and proposes two simultaneous optimization problems for Facial Attribute Classification.
Abstract: Recently, many researchers have attempted to classify Facial Attributes (FAs) by representing characteristics of FAs such as attractiveness, age, smiling and so on. In this context, recent studies have demonstrated that visual FAs are a strong background for many applications such as face verification, face search and so on. However, Facial Attribute Classification (FAC) in a wide range of attributes based on the regression representation -predicting of FAs as real-valued labels- is still a significant challenge in computer vision and psychology. In this paper, a regression model formulation is proposed for FAC in a wide range of FAs (e.g. 73 FAs). The proposed method accommodates real-valued scores to the probability of what percentage of the given FAs is present in the input image. To this end, two simultaneous dictionary learning methods are proposed to learn the regression and identity feature dictionaries simultaneously. Accordingly, a multi-level feature extraction is proposed for FAC. Then, four regression classification methods are proposed using a regression model formulated based on dictionary learning, SRC and CRC. Convincing results are acquired to handle a wide range of FAs and represent the probability of FAs on the PubFig, LFW, Groups and 10k US Adult Faces databases compared to several state-of-the-art methods. This paper proposes a method for regression facial attributes classification.We propose two simultaneous optimization problems for Facial Attribute Classification.A multilevel feature extraction method was proposed to discriminate the facial features.Promising results were obtained to classify facial attributes on the PubFig, Groups, 10k US adult and LFW databases.
TL;DR: A novel model, named Deep Attentive Multi-path Convolutional Neural Network (DAM-CNN), that can automatically locate expression-related regions in an expressional image and yield a robust image representation for FER.
Abstract: Facial Expression Recognition (FER) has long been a challenging task in the field of computer vision. In this paper, we present a novel model, named Deep Attentive Multi-path Convolutional Neural Network (DAM-CNN), for FER. Different from most existing models, DAM-CNN can automatically locate expression-related regions in an expressional image and yield a robust image representation for FER. The proposed model contains two novel modules: an attention-based Salient Expressional Region Descriptor (SERD) and the Multi-Path Variation-Suppressing Network (MPVS-Net). SERD can adaptively estimate the importance of different image regions for FER task, while MPVS-Net disentangles expressional information from irrelevant variations. By jointly combining SERD and MPVS-Net, DAM-CNN is able to highlight expression-relevant features and generate a variation-robust representation for expression classification. Extensive experimental results on both constrained datasets (CK+, JAFFE, TFEID) and unconstrained datasets (SFEW, FER2013, BAUM-2i) demonstrate the effectiveness of our DAM-CNN model.
TL;DR: An effective performance analysis of the proposed as well as the conventional methods such as convolutional neural network, NN-Levenberg-Marquardt, N nN-Gradient Descent, N N-Evolutionary Algorithm, Nn-firefly, and N n-Particle Swarm Optimisation is provided by evaluating few performance measures and thereby, the effectiveness of the suggested strategy over the conventional method is validated.
Abstract: The channels used to convey the human emotions consider actions, behaviours, poses, facial expressions, and speech. An immense research has been carried out to analyse the relationship between the facial emotions and these channels. The goal of this study is to develop a system for Facial Emotion Recognition (FER) that can analyse the elemental facial expressions of human, such as normal, smile, sad, surprise, anger, fear, and disgust. The recognition process of the proposed FER system is categorised into four processes, namely pre-processing, feature extraction, feature selection, and classification. After preprocessing, scale invariant feature transform -based feature extraction method is used to extract the features from the facial point. Further, a meta-heuristic algorithm called Grey Wolf optimisation (GWO) is used to select the optimal features. Subsequently, GWO-based neural network (NN) is used to classify the emotions from the selected features. Moreover, an effective performance analysis of the proposed as well as the conventional methods such as convolutional neural network, NN-Levenberg-Marquardt, NN-Gradient Descent, NN-Evolutionary Algorithm, NN-firefly, and NN-Particle Swarm Optimisation is provided by evaluating few performance measures and thereby, the effectiveness of the proposed strategy over the conventional methods is validated.
TL;DR: This work designs the state-of-the-art gender recognition and age estimation models according to three popular benchmarks: LFW, MORPH-II and FG-NET, significantly outperforming the solutions of other participants and winning the ChaLearn Apparent Age Estimation Challenge 2016.
Abstract: Convolutional Neural Networks (CNNs) have been proven very effective for human demographics estimation by a number of recent studies. However, the proposed solutions significantly vary in different aspects leaving many open questions on how to choose an optimal CNN architecture and which training strategy to use. In this work, we shed light on some of these questions improving the existing CNN-based approaches for gender and age prediction and providing practical hints for future studies. In particular, we analyse four important factors of the CNN training for gender recognition and age estimation: (1) the target age encoding and loss function, (2) the CNN depth, (3) the need for pretraining, and (4) the training strategy: mono-task or multi-task. As a result, we design the state-of-the-art gender recognition and age estimation models according to three popular benchmarks: LFW, MORPH-II and FG-NET. Moreover, our best model won the ChaLearn Apparent Age Estimation Challenge 2016 significantly outperforming the solutions of other participants.
TL;DR: A model based on Generative Adversarial Network (GAN) is proposed to address the open-set recognition without manual intervention during the training process, and it is shown that the proposed architecture outperforms other variants and is robust on both datasets.
Abstract: Open-set activity recognition remains as a challenging problem because of complex activity diversity. In previous works, extensive efforts have been paid to construct a negative set or set an optimal threshold for the target set. In this paper, a model based on Generative Adversarial Network (GAN), called ‘OpenGAN’ is proposed to address the open-set recognition without manual intervention during the training process. The generator produces fake target samples, which serve as an automatic negative set, and the discriminator is redesigned to output multiple categories together with an ‘unknown’ class. We evaluate the effectiveness of the proposed method on measured micro-Doppler radar dataset and the MOtion CAPture (MOCAP) database from Carnegie Mellon University (CMU). The comparison results with several state-of-the-art methods indicate that OpenGAN provides a promising open-set solution to human activity recognition even under the circumstance with few known classes. Ablation studies are also performed, and it is shown that the proposed architecture outperforms other variants and is robust on both datasets.
TL;DR: The obtained results show that multi-class data hyper plane using LDA and threefold SVM approach is effective and simple for quadratic data analysis.
Abstract: Representation and classification of multi-dimensional data are current key research areas. The representation of data in two classes is more feasible than multi-class representations because of the inherent quadratic complexity in existing techniques. Erroneous assignment of class labels affects separation boundary and training time complexity. In this paper, multi-dimensional data is handled using linear discriminant analysis (LDA) and threefold support vector machine (SVM) techniques to reduce the complexity and minimize false labeling. A facial expression application is proposed in which six natural expressions are used as multi-class data. Face image is divided into seven triangles on the basis of two focal points. A combined local and global feature descriptor is generated. Discrete Fourier transform is applied and processed with LDA to obtain discriminant features and accurately map an input feature space to an output space. To evaluate the system performance, Japanese Female Facial Expression, FER-2013 and Cohn–Kanade DFAT datasets are used. The obtained results show that multi-class data hyper plane using LDA and threefold SVM approach is effective and simple for quadratic data analysis