No abstract

Abstract-Bio-medical wearable devices restricted to their smallcapacity embedded-battery require energy-efficiency of the highest order. However, minimum-energy point (MEP) at sub-threshold voltages is unattainable with SRAM memory, which fails to hold below 0.3V because of its vanishing noise margins. This paper examines minimum-energy operation of 2T and 3T1D e-DRAM gain cells as an alternative to SRAM at 32nm technology node with different design points: up-sizing transistors, using high-Vth transistors, read/write wordline assists and temperature. First, the e-DRAM cells are evaluated without considering any process variations. The design-space is explored by creating a kriging meta-model to reduce the number of simulations. A full-factorial statistical analysis of e-DRAM cells is performed in presence of threshold voltage variations and the effect of upsizing on mean MEP is reported. Finally, it is shown that the product of the read and write lengths provides a knob for trade-off between energy-efficiency and reliable MEP energy operation.

Face recognition is one of the multimedia items that has seen a remarkable increase in popularity in recent years. Face continues to be the most difficult study topic for experts in the field of computer vision and image processing since it is an item with different properties for detection. We have attempted to handle the most challenging facial aspects in this survey work, including posture invariance, aging, illuminations, and partial occlusion. When applied to facial photographs, they are regarded as essential components of face recognition systems. The most recent face detection methods and techniques are also examined in this paper, including Eigenface, Artificial Neural Networks (ANN), Support Vector Machines (SVM), Principal Component Analysis (PCA), Independent Component Analysis (ICA), Gabor Wavelets, Elastic Bunch Graph Matching, 3D Morphable Models, and Hidden Markov Models. Many testing face databases, such as AT & T (ORL), AR, FERET, LFW, YTF, and Yale, also reviewed. However, the purpose of this study is to present a thorough literature assessment on face recognition and its applications.

With technology's increasing capabilities, social media has become the largest pool of data from which it can extract public opinion and begin to gather informative data on the success or failure of a brand, product, or marketing campaign in the eyes of the public. People share their experiences, opinions, and daily activities on social media, which results in enormous amounts of online data that attract developers to carry out data mining and analysis. Thus, there is a necessity for social media screening to obtain results that can be used for analysis. Twitter is an online networking site driven by tweets, which are 140-character limited messages. Thus, the character limit enforces the use of hashtags for text classification. Currently, around 5500–6000, tweets are published every second, which results in approximately 561.6 million tweets per day. Performing sentiment analysis of tweets can help us to determine the polarity and inclination of a vast population toward a specific topic, term, or entity. The applications of such analysis can easily be observed during public elections, movie promotions, brand endorsements, and many other fields. This proposed system uses a Naïve Bayes classifier to determine the tweets based on sentiment. In the implemented system, tweets are collected, and sentiment analysis is performed on them. Based on the sentiment analysis results, a few suggestions can be provided to the user. The primary aim is to provide a method for analyzing sentiment scores based on grades. This paper reports on the design of sentiment analysis, extracting vast numbers of tweets. Results classify users' perceptions via tweets into positive and negative categories. Secondly, it discusses various techniques to carry out a sentiment analysis on Twitter data in detail.

In this multilingual world, automatic detection of written or spoken language using Language Identification (LID) technology is a boon in the global communication with people using different languages in different countries. For simplicity and for the purpose of this research, the process of automatically identifying the language(s) from a document is thought of as LID. Lot of ongoing research projects are in the field of Natural Language Processing (NLP) that uses LID as a part of NLP. This field exploits several algorithms evolved in the field of computer science, individually or in combination to achieve accuracy in identifying a language. Among the different approaches adopted in LID,NaïveBayes Classification n-gram text processing seems to be promising.This paper proposes the concept for categorising multiple language texts using Naïve Bayesian algorithms using Machine Learning approaches. Using techniques from existing researches, this paper proposes a way to recognize multilingual documents and calculate the relative proportions of these languages.