Assessing the Performance of DWT-PCA/SVD Face Recognition Algorithm under Multiple Constraints

Abstract

Many architectures of face recognition modules have been developed to tackle the challenges posed by varying environmental constraints such as illumination, occlusions, pose, and expressions. These recognition systems have mainly focused on a single constraint at a time and have achieved remarkable successes. However, the presence of multiple constraints may deteriorate the performance of these face recognition systems. In this study, we assessed the performance of Principal Component Analysis and Singular Value Decomposition using Discrete Wavelet Transform (DWT-PCA/SVD) for preprocessing face recognition algorithm on multiple constraints (partially occluded face images acquired with varying expressions). Numerical evaluation of the study algorithm gave reasonably average recognition rates of 77.31% and 76.85% for left and right reconstructed face images with varying expressions, respectively. A statistically significant difference was established between the average recognition distance of the left and right reconstructed face images acquired with varying expressions using pairwise comparison test. The post hoc analysis using the Bonferroni simultaneous confidence interval revealed that the significant difference established through the pairwise comparison test was mainly due to the sad expressions. Although the performance of the DWT-PCA/SVD algorithm declined as compared to its performance on single constraints, the algorithm attained appreciable performance level under multiple constraints. The DWT-PCA/SVD recognition algorithm performs reasonably well for recognition when partial occlusion with varying expressions is the underlying constraint.

On the search for efficient face recognition algorithm subject to multiple environmental constraints

“…While the MICE imputation mechanism yielded quite impressive results (average recognition rates of 85.12% and 79.52% for 30% and 40% occlusion rates), the Missforest de-occlusion mechanism performs better at recovering missing portions of degraded face images acquired under expression variant constraint at moderately higher occlusion levels (40%). This result aligns with the findings of [25] and [10] .…”

“… [33] posited that the DWT presents a fast and easy way to filter facial images. [10] provides a thorough mathematical formulation of the DWT. This study examined the use of Haar wavelets to remove noise from a one-level, two-dimensional DWT.…”

“…This study examined the use of Haar wavelets to remove noise from a one-level, two-dimensional DWT. The Haar wavelet was chosen because of its computational simplicity and its orthogonal characteristic, which helps preserve distances after transformation [10] . A one-level DWT usually produces four sub-bands (LL, LH, HL, HH) of the image, where the LL sub-band denotes the approximate coefficients and the LH, HL and HH represents local information such as edges, noise, etc.…”

“…The performance of the PCA-based techniques is not yet optimum, especially on different constraints. [10] also assessed the performance of DWT-PCA/SVD algorithm on the constraint of partially occluded images acquired with varying facial expressions. The study achieved recognition rates of 77.31% and 76.85%, corresponding to the left and right reconstructed face images with variable facial expressions.…”