65 research outputs found
Improving Facial Emotion Recognition Using Residual Autoencoder Coupled Affinity Based Overlapping Reduction
No full textEmotion recognition using facial images has been a challenging task in computer vision. Recent advancements in deep learning has helped in achieving better results. Studies have pointed out that multiple facial expressions may present in facial images of a particular type of emotion. Thus, facial images of a category of emotion may have similarity to other categories of facial images, leading towards overlapping of classes in feature space. The problem of class overlapping has been studied primarily in the context of imbalanced classes. Few studies have considered imbalanced facial emotion recognition. However, to the authors’ best knowledge, no study has been found on the effects of overlapped classes on emotion recognition. Motivated by this, in the current study, an affinity-based overlap reduction technique (AFORET) has been proposed to deal with the overlapped class problem in facial emotion recognition. Firstly, a residual variational autoencoder (RVA) model has been used to transform the facial images to a latent vector form. Next, the proposed AFORET method has been applied on these overlapped latent vectors to reduce the overlapping between classes. The proposed method has been validated by training and testing various well known classifiers and comparing their performance in terms of a well known set of performance indicators. In addition, the proposed AFORET method is compared with already existing overlap reduction techniques, such as the OSM, ν-SVM, and NBU methods. Experimental results have shown that the proposed AFORET algorithm, when used with the RVA model, boosts classifier performance to a greater extent in predicting human emotion using facial images
Towards Golden Rule of Capital Accumulation: A Genetic Algorithm Approach
No full textPart 6: Modelling and OptimizationInternational audienceThe current study deals with maximizing consumption per worker in connection with the economic growth of society. The traditional Solow model based approach is well-studied and computationally complex. The present work proposes a Genetic Algorithm (GA) based consumption maximization in attaining the Golden rule. An objective function derived from traditional Solow model based on depreciation rate and amount of accumulated capital is utilized. The current study considered a constant output per worker to incorporate a constant efficiency level of labor. Different ranges of Depreciation rate and accumulated capital are tested to check the stability of the proposed GA based optimization process. The mean error and standard deviation in optimization process is utilized as a performance metric. The experimental results suggested that GA is very fast and is able to produce economically significant result with an average mean error 0.142% and standard deviation 0.021%
Non-Dominated Sorting Genetic Algorithm-II-Induced Neural-Supported Prediction of Water Quality with Stability Analysis
No full text Water is one of the most important necessities for human survival. In municipal corporation areas, water quality affects a large part of the population. Good quality water supply is an imperative parameter that influences individuals’ health. Automated accurate water quality determination becomes an urgent necessity. Detecting the drinking water quality can prevent such scenarios prior to the critical stage. Recent research works have achieved reasonable success in predicting the water quality by deploying several machine learning-based techniques and utilising different aspects to analyse water quality. The accuracy levels of already proposed models are to be improved, keeping in mind the sensitivity of the problem domain. In the current work, Non-dominated Sorting Genetic Algorithm-II (NN-NSGA-II) was employed to train the artificial neural network (ANN) to improve its performance over its traditional counterparts. The proposed model gradually minimises two different objective functions, namely the root mean square error (RMSE) and Maximum Error (ME) in order to find the optimal weight vector for the ANN. The proposed model was compared with another two well-established models namely ANN trained with Genetic Algorithm (NN-GA) and ANN trained with Particle Swarm Optimisation (NN-PSO) in terms of accuracy, precision, recall, [Formula: see text]-Measure, Matthews correlation coefficient (MCC) and Fowlkes–Mallows (FM) index. Furthermore, a data perturbation-based stability analysis is proposed to test the stability of the proposed method. The simulation results established superior accuracy of NN-NSGA-II over the other models. </jats:p
Effects of Class Imbalance Problem in Convolutional Neural Network Based Image Classification
No full text
- …
