Discovering oral cavity cancer (OCC) at an early stage is an effective way to increase patient survival rate. However, current initial screening process is done manually and is expensive for the average individual, especially in developing countries worldwide. This problem is further compounded due to the lack of specialists in such areas. Automating the initial screening process using artificial intelligence (AI) to detect pre-cancerous lesions can prove to be an effective and inexpensive technique that would allow patients to be triaged accordingly to receive appropriate clinical management. In this study, we have applied and evaluated the efficacy of six deep convolutional neural network (DCNN) models using transfer learning, for identifying pre-cancerous tongue lesions directly using a small dataset of clinically annotated photographic images to diagnose early signs of OCC. DCNN models were able to differentiate between benign and pre-cancerous tongue lesions and were also able to distinguish between five types of tongue lesions, i.e. hairy tongue, fissured tongue, geographic tongue, strawberry tongue and oral hairy leukoplakia with high classification performances. Preliminary results using an (AI + Physician) ensemble model demonstrate that an automated pre-screening process of oral tongue lesions using DCNNs can achieve ‘near-human’ level classification performance for diagnosing early signs of OCC in patients.

Early detection of brain tumors can save precious human life. This work presents a fully automated design to classify brain tumors. The proposed scheme employs optimal deep learning features for the classification of FLAIR, T1, T2, and T1CE tumors. Initially, we normalized the dataset to pass them to the ResNet101 pretrained model to perform transfer learning for our dataset. This approach results in fine-tuning the ResNet101 model for brain tumor classification. The problem with this approach is the generation of redundant features. These redundant features degrade accuracy and cause computational overhead. To tackle this problem, we find optimal features by utilizing differential evaluation and particle swarm optimization algorithms. The obtained optimal feature vectors are then serially fused to get a single-fused feature vector. PCA is applied to this fused vector to get the final optimized feature vector. This optimized feature vector is fed as input to various classifiers to classify tumors. Performance is analyzed at various stages. Performance results show that the proposed technique achieved a speedup of 25.5x in prediction time on the medium neural network with an accuracy of 94.4%. These results show significant improvement over the state-of-the-art techniques in terms of computational overhead by maintaining approximately the same accuracy.

Space modulation techniques (SMTs) have emerged as promising candidates for spectral- and energy-efficient wireless communication systems since they strike a good balance among error performance, power efficiency, spectrum efficiency, and receiver complexity. In SMTs, the information is not only conveyed by the habitual M-ary signal constellations; rather, it is also conveyed by the indices of the transmit antennas. As such, the indices of the transmit antennas are harnessed in such a manner that enhances the transmission efficiency compared to the other multiple-input multiple-output (MIMO) opponents. Despite their exceptional advantages, SMTs suffer from a major drawback, which lies in the logarithmic proportion between their achievable data rates and the number of transmit antennas. In this regard, the fully-generalised spatial modulation (F-GSM) and the fully-quadrature spatial modulation (F-QSM) are proposed in this paper in order to vanquish this controversial drawback. In F-GSM and F-QSM, the transmit antennas used for data transmission are varied from the state in which only one transmit antenna is activated to the state in which multiple/all transmit antennas are activated. Therefore, a linear relationship between the achievable data rates and the number of transmit antennas is acquired. Moreover, a novel mathematical framework for assessing the average bit error rate (ABER) performance of the different SMTs is delineated. The driven mathematical framework is considered as the first major attempt to generalize the analytical analysis of different SMTs. In addition, the receiver’s computational complexity of the proposed schemes is obtained and analysed in terms of the computational complexity of different SMTs. The simulation results substantiate the validity of the analytical analysis conducted throughout the paper, as they are very akin to the obtained analytical formulas.

Facial emotion recognition (FER) is an important research area in artificial intelligence (AI) and has many applications i.e., face authentication systems, e-learning, entertainment, deepfakes detection, etc. FER is still a challenging task due to more intra-class variations of emotions. Although prior deep learning methods have achieved good performance for FER. However, still there exists a need to develop efficient and effective FER systems robust to certain conditions i.e., variations in illumination, face angles, gender, race, background settings, and people belonging to diverse geographical regions. Moreover, a generalized model for the classification of human emotions is required to be implemented in computer systems so that they can interact with humans according to their emotions and improve their interaction. This work presents a novel light-weight Efficient-SwishNet model for emotion recognition that is robust towards the aforementioned conditions. We have introduced a low-cost, smooth unbounded above and bounded below Swish activation function in our model. Property of unboundedness helps to avoid saturation while smoothing helps in optimization and generalization of the model. Performance of the proposed model is evaluated on five diverse datasets including CK+, JAFFE, FER-2013, KDEF, and FERG datasets. We also performed a cross-corpora evaluation to show the generalizability of our model. The proposed model achieves a very high recognition rate for all datasets that prove the merit of the proposed framework for both the human facial images and stylized cartoon characters. Moreover, we conducted an ablation study with different variants of our model to prove its efficiency and effectiveness for emotions identification.

According to its various features, the solar photovoltaics (PV) system is realized as a significant promising energy source to cope with energy shortcomings and environmental impacts like contamination. Therefore, it is mandatory to estimate and predict the output power for prediction intervals to avoid any power outage or urgent disturbances in the utility grid. These are challenging tasks as the solar PV output power depends on the weather variables data such as temperature and solar radiation. In this article, the estimation and forecast of solar PV output power are investigated with an upscaling method using three different types of artificial neural networks (ANNs) in order to reduce the estimation errors in current types of ANNs. The multilayer feedforward neural network (MLFFNN), recurrent neural network (RNN), and nonlinear autoregressive exogenous (NARX) model neural network (NARXNN) are applied to estimate and forecast the total output power of four real solar PV substations in Egypt. Hence both the surface temperature and the solar radiation of each PV substation are applied as the inputs of each designed NN, whereas the total output power of the four PV substations is its output. For the training and effectiveness investigation procedures of each applied ANNs, the data of two months (60 days) are attained and collected from these four PV substations. Here, the data of the first 45 days are applied to train the three designed NNs, while the data from the remaining 15 days, which are not applied for the training, are used to check the effectiveness and the generalization capability of the trained NNs. Hence, the estimation process is considered a prior step for the forecast of the output power. Therefore, upscaling method is utilized for assessing and forecasting a regional solar PV output power because of the limited number of monitored plants and applied data. The results provide evidence that the trained NNs are running very well and efficiently to estimate the power correctly. The performance of the MLFFNN is the best compared with the other NNs, whereas the NARXNN’s performance is the lowest one. The MLFFNN achieves the lowest mean squared error (MSE) of 0.27533 and the lowest absolute approximation error of 0.2099 MWh. Finally, the assessment and comparison among the three trained NNs and other techniques in recently published articles are highlighted and presented which reveal the performance superiority of the three trained NNs compared to other ANNs.