Discriminant Correlation Analysis: Real-Time Feature Level Fusion for Multimodal Biometric Recognition

Abstract

Information fusion is a key step in multimodal biometric systems. Fusion of information can occur at different levels of a recognition system, i.e., at the feature level, matching-score level, or decision level. However, feature level fusion is believed to be more effective owing to the fact that a feature set contains richer information about the input biometric data than the matching score or the output decision of a classifier. The goal of feature fusion for recognition is to combine relevant information from two or more feature vectors into a single one with more discriminative power than any of the input feature vectors. In pattern recognition problems, we are also interested in separating the classes. In this paper, we present Discriminant Correlation Analysis (DCA), a feature level fusion technique that incorporates the class associations into the correlation analysis of the feature sets. DCA performs an effective feature fusion by maximizing the pairwise correlations across the two feature sets, and at the same time, eliminating the between-class correlations and restricting the correlations to be within the classes. Our proposed method can be used in pattern recognition applications for fusing features extracted from multiple modalities or combining different feature vectors extracted from a single modality. It is noteworthy that DCA is the first technique that considers class structure in feature fusion. Moreover, it has a very low computational complexity and it can be employed in real-time applications. Multiple sets of experiments performed on various biometric databases, and using different feature extraction techniques, show the effectiveness of our proposed method, which outperforms other state-of-the-art approaches.

Multi-Level Feature Abstraction from Convolutional Neural Networks for Multimodal Biometric Identification

“…The feature level fusion techniques include the serial feature fusion [17], the parallel feature fusion [18], the CCAbased feature fusion [19], JSRC [14], SMDL [20] and DCA/MDCA [16] algorithms. Note that in case of more than two modalities, the parallel feature fusion method cannot be applied.…”

“…Compared to score, rank, and decision level fusions, feature level fusion results in a better discriminative classifier [12], [13], due to preservation of raw information [14]. The feature level fusion integrates different features extracted from different modalities into a more abstract and compact feature representation, which can be further used for classification, ver-ification, or identification [15], [16]. Recently several authors have exploited feature level fusion for multimodal biometric identification.…”

A Method for Estimating the Predictive Power of a Binary Indicator

“…В частности, когда классы объектов, подлежащих изучению, долж-ным образом определены и описаны, эта проблема сводится к выбору того из имеющихся классов, к которому должен быть отнесен вновь обнару-женный объект, и получила название дискрими-нантного анализа, методы которого сегодня ак-тивно развиваются (см. [3]). Одно из востребо-ванных применений этого аппарата -задачи так называемой доказательной медицины, основы ко-торой изложены в [4,5].…”

A Short Review of Multimodal Biometric Recognition Systems

“…Haghighat et al [2] reported that feature level fusion can be more effective in terms of biometric data classification by owing to that a feature set which contains richer information in regard the input biometric data than that of matching score or decision level fusion. They proposed a discriminant correlation analysis (DCA), a new technique of feature level fusion that integrates the class associations into the correlation analysis of the feature sets.…”