Content uploaded by Madhuri Bhalekar
Author content
All content in this area was uploaded by Madhuri Bhalekar on Jan 04, 2022
Content may be subject to copyright.
A preview of the PDF is not available
A Model for Automated Food Logging Through Food Recognition and Attribute Estimation Using Deep Learning
Abstract
The past few decades have witnessed an increase in dietary ailments, majorly caused due to unhealthy food habits. According to experts, mobile-based diet monitoring and assessment systems can capture real-time images of various food items as well as analyze the nutritional content present in it and can be very convenient to use and assist in improving food habits. This can help people lead a healthier lifestyle. This proposed model provides an innovative system that can automatically estimate various food attributes like the nutrients and ingredients by classifying the food image that is given as input. The approach involves the use of different types of deep learning models for accurate food item identification. Apart from image recognition and analysis, the food ingredients, nutrients and attributes are obtained and estimated by extracting words which are semantically related from a large collection of text, accumulated over the Internet. Experimentation has been performed with the Food-101 dataset. The proposed system assists the user to obtain the nutritional value of the food item in real-time which is effective and simple to use. The proposed system also provides supporting features such as food logging, calorie tracking and healthy recipe recommendations for self-monitoring of the user.
... Certain approaches provide nutritional information based on the average portion that a person could take or the average information could be in a meal [7], while others estimate the exact quantity of food item from an image [8]. Food recognition approaches have been adopted of generated several food data sets to evaluate their contribution and trait either a specific type of cuisine, such as Asian food [9] or miscellaneous types [10]. While others choose to recognize fruits and vegetables [11]. ...
... The authors used Tensorflow Lite to transfer the model to a mobile device to facilitate access to users. Ambadkar et al. [10] implemented a system to recognize food images and estimate attributes, ingredients and nutritional information for web and mobile use. For food recognition, authors used four pretrained models: inception-v3, inception-v4, Xception and inceptionRestNetV2. ...
The management of daily food intake aids to preserve a healthy body, minimize the risk of many diseases, and monitor chronic diseases, such as diabetes and heart problems. To ensure a healthy food intake, artificial intelligence has been widely used for food image recognition and nutrition analysis. Several approaches have been generated using a powerful type of machine learning: deep learning. In this paper, a systematic review is presented for the application of deep learning in food image recognition and nutrition analysis. A methodology of systematic research has been adopted resulting in three main fields: food image classification, food image segmentation and volume estimation of food items providing nutritional information ."57" original articles were selected and synthesized based on the use case of the approach, the employed model, the used data set, the experiment process and finally the main results. In addition, articles of public and private food data sets are presented. It is noted that among the literature review, several deep learning-based studies have shown great results and outperform the conventional methods. However, certain challenges and limitations are presented. Hence, some research directions are proposed to apply in the future to improve the food recognition systems for dietary assessment.
Quantum computers use the superposition of quantum bits or qubit’s in
the two quantum states, which are |0> and |1> and work faster than classical computers.
The qubits in the quantum computer are the basic unit of quantum information and replacement of the traditional bits 0 and 1 of classical computing. The n qubits can be simultaneously represented at 2n states based on quantum theory. For example, two
qubits can be depicted in four states which are 00, 01, 10, or 11 at the same time due to the superposition phenomenon of atoms. The real advantage of quantum theory can
be recognized by adding more qubits that help to grow the power of the computer exponentially.
Due to the rapid growth of the Internet, users change their preference from traditional shopping to the electronic commerce. Instead of bank/shop robbery, nowadays, criminals try to find their victims in the cyberspace with some specific tricks. By using the anonymous structure of the Internet, attackers set out new techniques, such as phishing, to deceive victims with the use of false websites to collect their sensitive information such as account IDs, usernames, passwords, etc. Understanding whether a web page is legitimate or phishing is a very challenging problem, due to its semantics-based attack structure , which mainly exploits the computer users' vulnerabilities. Although software companies launch new anti-phishing products, which use blacklists, heuristics, visual and machine learning-based approaches, these products cannot prevent all of the phishing attacks. In this paper, a real-time anti-phishing system, which uses seven different classification algorithms and natural language processing (NLP) based features, is proposed. The system has the following distinguishing properties from other studies in the literature: language independence, use of a huge size of phishing and legitimate data, real-time execution, detection of new websites, independence from third-party services and use of feature-rich classifiers. For measuring the performance of the system, a new dataset is constructed, and the experimental results are tested on it. According to the experimental and comparative results from the implemented classification algorithms, Random Forest algorithm with only NLP based features gives the best performance with the 97.98% accuracy rate for detection of phishing URLs.
We report on the results of an ongoing, long-term, large-scale, cross-organizational, and canonical action research project that we started in Estonia in 2003 to develop a reusable model (e-LocGov model) for the systematic introduction of e-governance into local governments. The project was conducted in cooperation with 126 local governments, which constitute more than half of Estonia's local governments. Given the complexity of the task, the endeavors in this project are necessarily highly interdisciplinary and cover the legal, the managerial, and the technological spheres. Consequentially, the e-LocGov model consists of major building blocks, each of which meets with a major challenge in the introduction of e-Governance: (i) state readiness, (ii) local government readiness, (iii) transition/implementation, and (iv) analysis and assessment. We provide a comprehensive explanation of the e-LocGov model and its building blocks. Also, we report on the research methodology and the theoretical background of this project.
As the amount of data available exponentially grows, data scientists are aware that finding the value in the data is key to a successful data exploiting. However, the data rarely presents itself in a ordered, clean way. In opposition to dealing with raw data, the term smart data is becoming more and more visible both in the specialized literature and companies. While software packages publicly exist to deal with raw data, there is no unified framework that encompasses all the required fields to transform such raw data to smart data. In this paper, the novel smartdata package is introduced. Written in R and available at CRAN repository, it includes the most recent and relevant algorithms to treat raw data from multiple perspectives, now unified under a simple yet powerful API, which enables the data scientist to easily pipeline their application. The main features of the package, as well as some illustrative examples of its use are detailed throughout this manuscript.
