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

From literature, majority of face recognition modules suffer performance challenges when presented with test images acquired under multiple constrained environments (occlusion and varying expressions). The performance of these models further deteriorates as the degree of degradation of the test images increases (relatively higher occlusion level). Deep learning-based face recognition models have attracted much attention in the research community as they are purported to outperform the classical PCA-based methods. Unfortunately their application to real-life problems is limited because of their intensive computational complexity and relatively longer run-times. This study proposes an enhancement of some PCA-based methods (with relatively lower computational complexity and run-time) to overcome the challenges posed to the recognition module in the presence of multiple constraints. The study compared the performance of enhanced classical PCA-based method (HE-GC-DWT-PCA/SVD) to FaceNet algorithm (deep learning method) using expression variant face images artificially occluded at 30% and 40%. The study leveraged on two statistical imputation methods of MissForest and Multiple Imputation by Chained Equations (MICE) for occlusion recovery. From the numerical evaluation results, although the two models achieved the same recognition rate (85.19%) at 30% level of occlusion, the enhanced PCA-based algorithm (HE-GC-DWT-PCA/SVD) outperformed the FaceNet model at 40% occlusion rate, with a recognition rate of 83.33%. Although both Missforest and MICE performed creditably well as de-occlusion mechanisms at higher levels of occlusion, MissForest outperforms the MICE imputation mechanism. MissForest imputation mechanism and the proposed HE-GC-DWT-PCA/SVD algorithm are recommended for occlusion recovery and recognition of multiple constrained test images respectively.

Time series based road traffic accidents forecasting via SARIMA and Facebook Prophet model with potential changepoints.

Road traffic accident (RTA) is a critical global public health concern, particularly in developing countries. Analyzing past fatalities and predicting future trends is vital for the development of road safety policies and regulations. The main objective of this study is to assess the effectiveness of univariate Seasonal Autoregressive Integrated Moving Average (SARIMA) and Facebook (FB) Prophet models, with potential change points, in handling time-series road accident data involving seasonal patterns in contrast to other statistical methods employed by key governmental agencies such as Ghana's Motor Transport and Traffic Unit (MTTU). The aforementioned models underwent training with monthly RTA data spanning from 2013 to 2018. Their predictive accuracies were then evaluated using the test set, comprising monthly RTA data from 2019. The study employed the Box-Jenkins method on the training set, yielding the development of various tentative time series models to effectively capture the patterns in the monthly RTA data. SARIMA(0,1,1)×(1,0,0)12 was found to be the suitable model for forecasting RTAs with a log-likelihood value of -266.28, AIC value of 538.56, AICc value of 538.92, BIC value of 545.35. The findings disclosed that the SARIMA(0,1,1)×(1,0,0)12 model developed outperforms FB-Prophet with a forecast accuracy of 93.1025% as clearly depicted by the model's MAPE of 6.8975% and a Theil U1 statistic of 0.0376 compared to the FB-Prophet model's respective forecasted accuracy and Theil U1 statistic of 84.3569% and 0.1071. A Ljung-Box test on the residuals of the estimated SARIMA(0,1,1)×(1,0,0)12 model revealed that they are independent and free from auto/serial correlation. A Box-Pierce test for larger lags also revealed that the proposed model is adequate for forecasting. Due to the high forecast accuracy of the proposed SARIMA model, the study recommends the use of the proposed SARIMA model in the analysis of road traffic accidents in Ghana.