Giovani Chiachia

Deep Representations for Iris, Face, and Fingerprint Spoofing Detection

Biometrics systems have significantly improved person identification and authentication, playing an important role in personal, national, and global security. However, these systems might be deceived (or spoofed) and, despite the recent advances in spoofing detection, current solutions often rely on domain knowledge, specific biometric reading systems, and attack types. We assume a very limited knowledge about biometric spoofing at the sensor to derive outstanding spoofing detection systems for iris, face, and fingerprint modalities based on two deep learning approaches. The first approach consists of learning suitable convolutional network architectures for each domain, whereas the second approach focuses on learning the weights of the network via back propagation. We consider nine biometric spoofing benchmarks - each one containing real and fake samples of a given biometric modality and attack type - and learn deep representations for each benchmark by combining and contrasting the two learning approaches. This strategy not only provides better comprehension of how these approaches interplay, but also creates systems that exceed the best known results in eight out of the nine benchmarks. The results strongly indicate that spoofing detection systems based on convolutional networks can be robust to attacks already known and possibly adapted, with little effort, to image-based attacks that are yet to come.

A novel algorithm for feature selection using Harmony Search and its application for non-technical losses detection

Finding an optimal subset of features that maximizes classification accuracy is still an open problem. In this paper, we exploit the speed of the Harmony Search algorithm and the Optimum-Path Forest classifier in order to propose a new fast and accurate approach for feature selection. Comparisons to some other pattern recognition and feature selection techniques showed that the proposed hybrid algorithm for feature selection outperformed them. The experiments were carried out in the context of identifying non-technical losses in power distribution systems.

Feature selection through gravitational search algorithm

In this paper we deal with the problem of feature selection by introducing a new approach based on Gravitational Search Algorithm (GSA). The proposed algorithm combines the optimization behavior of GSA together with the speed of Optimum-Path Forest (OPF) classifier in order to provide a fast and accurate framework for feature selection. Experiments on datasets obtained from a wide range of applications, such as vowel recognition, image classification and fraud detection in power distribution systems are conducted in order to asses the robustness of the proposed technique against Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and a Particle Swarm Optimization (PSO)-based algorithm for feature selection.

Learning Person-Specific Representations From Faces in the Wild

Humans are natural face recognition experts, far out-performing current automated face recognition algorithms, especially in naturalistic, “in the wild” settings. However, a striking feature of human face recognition is that we are dramatically better at recognizing highly familiar faces, presumably because we can leverage large amounts of past experience with the appearance of an individual to aid future recognition. Meanwhile, the analogous situation in automated face recognition, where a large number of training examples of an individual are available, has been largely underexplored, in spite of the increasing relevance of this setting in the age of social media. Inspired by these observations, we propose to explicitly learn enhanced face representations on a per-individual basis, and we present two methods enabling this approach. By learning and operating within person-specific representations, we are able to significantly outperform the previous state-of-the-art on PubFig83, a challenging benchmark for familiar face recognition in the wild, using a novel method for learning representations in deep visual hierarchies. We suggest that such person-specific representations aid recognition by introducing an intermediate form of regularization to the problem.

Vehicle License Plate Recognition With Random Convolutional Networks

Despite decades of research on automatic license plate recognition (ALPR), optical character recognition (OCR) still leaves room for improvement in this context, given that a single OCR miss is enough to miss the entire plate. We propose an OCR approach based on convolutional neural networks (CNNs) for feature extraction. The architecture of our CNN is chosen from thousands of random possibilities and its filter weights are set at random and normalized to zero mean and unit norm. By training linear support vector machines (SVMs) on the resulting CNN features, we can achieve recognition rates of over 98% for digits and 96% for letters, something that neither SVMs operating on image pixels nor CNNs trained via back-propagation can achieve. The results are obtained in a dataset that has 182 samples per digit and 28 per letter, and suggest the use of random CNNs as a promising alternative approach to ALPR systems.

Person-Specific Subspace Analysis for Unconstrained Familiar Face Identification.

While significant strides have been made in the recognition of faces under controlled viewing conditions, face recognition “in the wild” remains a challenging unsolved problem [12, 13, 21]. Interestingly, while humans are generally excellent at identifying familiar individuals under such conditions, their performance is significantly worse with unfamiliar individuals [5] and groups [19], leading to the idea that brain may have enhanced or specialized representations of familiar individuals [6]. Inspired by these observations, we explored the use of a number of subspace analysis techniques, applied to various visual representations, to generate person-specific subspaces of “familiar” individuals for face identification. In particular, we introduce a person-specific application of partial least squares (PS-PLS) to generate per-individual subspaces, and show that operating in these subspaces yields state-of-the-art performance on the challenging PubFig83 familiar face identification benchmark. The results underscore the potential importance of incorporating a notion of familiarity into face recognition systems.

What is the importance of selecting features for non-technical losses identification?

Although non-technical losses automatic identification has been massively studied, the problem of selecting the most representative features in order to boost the identification accuracy has not attracted much attention in this context. In this paper, we focus on this problem applying a novel feature selection algorithm based on Particle Swarm Optimization and Optimum-Path Forest. The results demonstrated that this method can improve the classification accuracy of possible frauds up to 49% in some datasets composed by industrial and commercial profiles.

Petroleum well drilling monitoring through cutting image analysis and artificial intelligence techniques

Petroleum well drilling monitoring has become an important tool for detecting and preventing problems during the well drilling process. In this paper, we propose to assist the drilling process by analyzing the cutting images at the vibrating shake shaker, in which different concentrations of cuttings can indicate possible problems, such as the collapse of the well borehole walls. In such a way, we present here an innovative computer vision system composed by a real time cutting volume estimator addressed by support vector regression. As far we know, we are the first to propose the petroleum well drilling monitoring by cutting image analysis. We also applied a collection of supervised classifiers for cutting volume classification.

CENSUS HISTOGRAMS: A SIMPLE FEATURE EXTRACTION AND MATCHING APPROACH FOR FACE RECOGNITION

Most face recognition approaches require a prior training where a given distribution of faces is assumed to further predict the identity of test faces. Such an approach may experience difficulty in...

Infrared Face Recognition by Optimum-Path Forest

This paper presents a novel, fast and accurate appearance-based method for infrared face recognition. By introducing the Optimum-Path Forest classifier, our objective is to get good recognition rates and effectively reduce the computational effort. The feature extraction procedure is carried out by PCA, and the results are compared to two other well known supervised learning classifiers; Artificial Neural Networks and Support Vector Machines. The achieved performance asserts the promise of the proposed framework.