Periapical dental X-ray image classification using deep neural networks.

This paper studies the problem of detection of dental diseases. Dental problems affect the vast majority of the world's population. Caries, RCT (Root Canal Treatment), Abscess, Bone Loss, and missing teeth are some of the most common dental conditions that affect people of all ages all over the world. Delayed or incorrect diagnosis may result in mistreatment, affecting not only an individual's oral health but also his or her overall health, thereby making it an important research area in medicine and engineering. We propose a pipelined Deep Neural Network (DNN) approach to detect healthy and non-healthy periapical dental X-ray images. Even a minor enhancement or improvement in existing techniques can go a long way in providing significant health benefits in the medical field. This paper has made a successful attempt to contribute a different type of pipelined approach using AlexNet in this regard. The approach is trained on a large dataset of 16,000 dental X-ray images, correctly identifying healthy and non-healthy X-ray images. We use an optimized Convolutional Neural Networks and three state-of-the-art DNN models, namely Res-Net-18, ResNet-34, and AlexNet for disease classification. In our study, the AlexNet model outperforms the other models with an accuracy of 0.852. The precision, recall and F1 scores of AlexNet also surpass the other models with a score of 0.850 across all metrics. The area under ROC curve also signifies that both the false-positive rate and false-negative rate are low. We conclude that even with a big data set and raw X-ray pictures, the AlexNet model generalizes effectively to previously unseen data and can aid in the diagnosis of a variety of dental diseases.

Understanding cartoon emotion using integrated deep neural network on large dataset.

Emotion is an instinctive or intuitive feeling as distinguished from reasoning or knowledge. It varies over time, since it is a natural instinctive state of mind deriving from one's circumstances, mood, or relationships with others. Since emotions vary over time, it is important to understand and analyze them appropriately. Existing works have mostly focused well on recognizing basic emotions from human faces. However, the emotion recognition from cartoon images has not been extensively covered. Therefore, in this paper, we present an integrated Deep Neural Network (DNN) approach that deals with recognizing emotions from cartoon images. Since state-of-works do not have large amount of data, we collected a dataset of size 8 K from two cartoon characters: 'Tom' & 'Jerry' with four different emotions, namely happy, sad, angry, and surprise. The proposed integrated DNN approach, trained on a large dataset consisting of animations for both the characters (Tom and Jerry), correctly identifies the character, segments their face masks, and recognizes the consequent emotions with an accuracy score of 0.96. The approach utilizes Mask R-CNN for character detection and state-of-the-art deep learning models, namely ResNet-50, MobileNetV2, InceptionV3, and VGG 16 for emotion classification. In our study, to classify emotions, VGG 16 outperforms others with an accuracy of 96% and F1 score of 0.85. The proposed integrated DNN outperforms the state-of-the-art approaches.

Production of biofuels from biomass: Predicting the energy employing artificial intelligence modelling.

Bioenergy may be a major replacement of fossil fuels which can make the path easier for sustainable development and decrease the dependency on conventional sources of energy. The main concern with the bioenergy is the availability of feedstock, dealing with its economics as well as its demand and supply chain management. This review deals with the finding of distinct potential of different Artificial Intelligence technologies focusing the challenges in bioenergy production system and its overall improvement in application. The study also highlights the contribution of Artificial Intelligence techniques for the prediction of energy from biomass and evaluates the computing-reasoning techniques for managing bioenergy production, biomass supply chain and optimization of process parameters for efficient bioconversion technologies.

A Novel MEMS Capacitive Microphone with Semiconstrained Diaphragm Supported with Center and Peripheral Backplate Protrusions.

Audio applications such as mobile phones, hearing aids, true wireless stereo earphones, and Internet of Things devices demand small size, high performance, and reduced cost. Microelectromechanical system (MEMS) capacitive microphones fulfill these requirements with improved reliability and specifications related to sensitivity, signal-to-noise ratio (SNR), distortion, and dynamic range when compared to their electret condenser microphone counterparts. We present the design and modeling of a semiconstrained polysilicon diaphragm with flexible springs that are simply supported under bias voltage with a center and eight peripheral protrusions extending from the backplate. The flexible springs attached to the diaphragm reduce the residual film stress effect more effectively compared to constrained diaphragms. The center and peripheral protrusions from the backplate further increase the effective area, linearity, and sensitivity of the diaphragm when the diaphragm engages with these protrusions under an applied bias voltage. Finite element modeling approaches have been implemented to estimate deflection, compliance, and resonance. We report an 85% increase in the effective area of the diaphragm in this configuration with respect to a constrained diaphragm and a 48% increase with respect to a simply supported diaphragm without the center protrusion. Under the applied bias, the effective area further increases by an additional 15% as compared to the unbiased diaphragm effective area. A lumped element model has been also developed to predict the mechanical and electrical behavior of the microphone. With an applied bias, the microphone has a sensitivity of -38 dB (ref. 1 V/Pa at 1 kHz) and an SNR of 67 dBA measured in a 3.25 mm × 1.9 mm × 0.9 mm package including an analog ASIC.