Computer Science and Engineering

<p>Shiv Nadar University is a comprehensive, multidisciplinary, research-focused, and student-centric University offering a full range of academic programs at the undergraduate, postgraduate and doctoral level. With state-of-the-art infrastructure, the University comprises of academic wings, sophisticated labs, international standard sports facilities, amphitheaters, auditorium, conference rooms and smart classrooms. The University&rsquo;s goal is to become internationally recognized for the quality of its research and creative endeavors and their applicability to improving quality of life, generating new insights and expanding the boundaries of human knowledge creativity. Committed to excellence in teaching, research and service, the University aims to serve the higher education needs of India and the world beyond.</p>

Shiv Nadar University

Mobile application based ride-hailing systems, e.g., DiDi, Uber have become part of day to day life and natural choices of transport for urban commuters. However, the pick-up demand in any area is not always matching with the supply or drop-off request in the same area. Urban planners and researchers are working hard to balance this demand and supply situation for taxi requests. The existing approaches have mainly focused on clustering of the spatial regions to identify hotspots, which refer to the locations with a high demand for pick-up requests. In our study, we determined that if the hotspots focus on the clustering of high demand for pick-up requests, most of the hotspots pivot near the city center or two-three spatial regions, ignoring the other parts of the city. In this work, we proposed a method, which can help in finding a local hotspot to cover the whole city area. We proposed a dominating set problem based solution, which covers every part of the city. This will help the drivers looking for near-by next customer in the region wherever they drop their last customer. It will also reduce the waiting time for customers as well as for a driver looking for next pick-up request. This would maximize their profit as well as help in improving their services. © 2021, Springer Nature Switzerland AG.

(SoE) School of Engineering

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Optimal Placement of Taxis in a City Using Dominating Set Problem

Cyberbullying is one of the leading causes of mental health issues in the younger population, often leading to depression, stress, and suicidal tendencies. Often it is observed that timely intervention can bring awareness towards unintentional cyberbullying. This paper presents an approach to the prevention of cyberbullying via social networking. The objective of this work is real-time detection of the degree of offensiveness and alerting the user typing the message. We also propose corrective measures, by displaying an alternative word for any offensive word that is typed in the suggestion box in real-time. This enables the user to prevent unintentional cyberbullying. The proposed solution is integrated into an app, that displays the relevant statistics as well as visualization of the user typing pattern. The model to detect the offensiveness percentage can achieve 97.77\% accuracy and is the backbone of the entire approach. © 2021, Springer Nature Switzerland AG.

A Real Time Analysis of Offensive Texts to Prevent Cyberbullying

Diabetic Macular Edema (DME) is an advanced stage of Diabetic Retinopathy (DR) and can lead to permanent vision loss. Currently, it affects 26.7 million people globally and on account of such a huge number of DME cases and the limited number of ophthalmologists, it is desirable to automate the diagnosis process. Computer-assisted, deep learning based diagnosis could help in early detection, following which precision medication can help to mitigate the vision loss. Method: In order to automate the screening of DME, we propose a novel DMENet Algorithm which is built on the pillars of Convolutional Neural Networks (CNNs). DMENet analyses the preprocessed color fundus images and passes it through a two-stage pipeline. The first stage detects the presence or absence of DME whereas the second stage takes only the positive cases and grades the images based on severity. In both the stages, we use a novel Hierarchical Ensemble of CNNs (HE-CNN). This paper uses two of the popular publicly available datasets IDRiD and MESSIDOR for classification. Preprocessing on the images is performed using morphological opening and gaussian kernel. The dataset is augmented to solve the class imbalance problem for better performance of the proposed model. Results: The proposed methodology achieved an average Accuracy of 96.12\%, Sensitivity of 96.32\%, Specificity of 95.84\%, and F−1 score of 0.9609 on MESSIDOR and IDRiD datasets. Conclusion: These excellent results establish the validity of the proposed methodology for use in DME screening and solidifies the applicability of the HE-CNN classification technique in the domain of biomedical imaging. © 2020 Singh, Gorantla. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Fulltext

PLoS ONE

DMenet: Diabetic Macular Edema diagnosis using Hierarchical Ensemble of CNNs

Indoor occupancy estimation has become an important area of research in the recent past. This work investigates the feasibility of an Internet of Things (IoT) based university classroom occupancy estimation system. As IoT devices generate voluminous data at high rates, the centralized cloud computing approach is found to generate high latencies. The client server based cloud architecture has been compared with the decentralized edge computing architecture for building the occupancy estimation system. The performance of these architectures has been compared using two performance metrics: latency and throughput. The occupancy estimation models using carbon dioxide and relative humidity as inputs, have been developed using multiple linear regression and quantile regression. Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE) have been used to compare the performance of our estimation models. © 2020 The Authors. Published by Elsevier B.V.

Procedia Computer Science

IoT-based Indoor Occupancy Estimation Using Edge Computing

Good air quality is critical to healthy and productive life in the indoor environment. Quality of air inside buildings depends on several factors. Till now, it was not possible to measure and monitor indoor air quality (IAQ) through mobile phones and wireless sensors in real time, as standardized miniature sensors were not commercially available. With the introduction of wireless air quality sensors that can pair with most contemporary smartphones, it is now possible to monitor indoor environment pervasively in real time, take precautionary measures, and maintain a healthy life. In this work, we have created an IoT sensing system to monitor and analyse the time variation of carbon dioxide in a university classroom. Using the sensed data, forecast models have been developed to predict the build-up of carbon dioxide, an indicator of IAQ. © Springer Nature Singapore Pte Ltd. 2020.

Communications in Computer and Information Science

An Internet of Things Framework to Forecast Indoor Air Quality Using Machine Learning

In spite of the recent advancements of deep learning based techniques, automatic photo aesthetic assessment still remains a challenging computer vision task. Existing approaches used to focus on providing a single aesthetic score or category ("good" or "bad") of photograph, rather than quantifying "goodness" or "badness". The existing algorithms often ignore the importance of different attributes contributing to the artistic quality of the photograph. To obtain the human-interpretability of aesthetic score of photo, we advocate learning the aesthetic attributes alongwith the prediction of the general aesthetic score. We propose a multi-task deep CNN, that collectively learns aesthetic attributes alongwith a general aesthetic score for the photograph. To understand the mathematical representation of the attributes in the proposed model, a visualization technique is proposed using back propagation of gradients. These visualization of attributes correspond to the location of objects in the images in order to find out which part of an image "triggers" the classification outcome, thus providing the insights about the model's understanding of these attributes. This paper proposes an aesthetic feature vector based on the relative foreground position of the object in the image. The proposed aesthetic features outperform the state-of-art methods especially for Rule of Thirds attribute. © 2020 The Institution of Engineering and Technology.

World University root node

IET Image Processing

Measuring photography aesthetics with deep CNNs

Human Activity Recognition in RGB-D videos has been an active research topic during the last decade. However, only a few efforts have been made, for recognizing human activity in RGB-D videos where several performers are performing simultaneously. In this paper we introduce such a challenging dataset with several performers performing the activities simultaniously. We present a novel method for recognizing human activities performed simultaniously in the same videos. The proposed method aims in capturing the motion information of the whole video by producing a dynamic image corresponding to the input video. We use two parallel ResNet-101 architectures to produce the dynamic images for the RGB video and depth video separately. The dynamic images contain only the motion information of the whole frame, which is the main cue for analyzing the motion of the performer during action. Hence, dynamic images help recognizing human action by concentrating only on the motion information appeared on the frame. We send the two dynamic images through a fully connected layer for classification of activity. The proposed dynamic image reduces the complexity of the recognition process by extracting a sparse matrix from a video, while preserving the motion information required for activity recognition, and produces comparable results with respect to the state-of-the-art. © 2020, Springer Science+Business Media, LLC, part of Springer Nature.

Multimedia Tools and Applications

Human activity recognition in RGB-D videos by dynamic images

Stochastic gradient descent (SGD) is one of the core techniques behind the success of deep neural networks. The gradient provides information on the direction in which a function has the steepest rate of change. The main problem with basic SGD is to change by equal-sized steps for all parameters, irrespective of the gradient behavior. Hence, an efficient way of deep network optimization is to have adaptive step sizes for each parameter. Recently, several attempts have been made to improve gradient descent methods such as AdaGrad, AdaDelta, RMSProp, and adaptive moment estimation (Adam). These methods rely on the square roots of exponential moving averages of squared past gradients. Thus, these methods do not take advantage of local change in gradients. In this article, a novel optimizer is proposed based on the difference between the present and the immediate past gradient (i.e., diffGrad). In the proposed diffGrad optimization technique, the step size is adjusted for each parameter in such a way that it should have a larger step size for faster gradient changing parameters and a lower step size for lower gradient changing parameters. The convergence analysis is done using the regret bound approach of the online learning framework. In this article, thorough analysis is made over three synthetic complex nonconvex functions. The image categorization experiments are also conducted over the CIFAR10 and CIFAR100 data sets to observe the performance of diffGrad with respect to the state-of-the-art optimizers such as SGDM, AdaGrad, AdaDelta, RMSProp, AMSGrad, and Adam. The residual unit (ResNet)-based convolutional neural network (CNN) architecture is used in the experiments. The experiments show that diffGrad outperforms other optimizers. Also, we show that diffGrad performs uniformly well for training CNN using different activation functions. The source code is made publicly available at https://github.com/shivram1987/diffGrad. © 2012 IEEE.

IEEE Transactions on Neural Networks and Learning Systems

DiffGrad: An Optimization Method for Convolutional Neural Networks

The human attention span is continuously decreasing, and the amount of time a person wants to spend on reading is declining at an alarming rate. Therefore, it is imperative to provide a quick glance of important news by generating a concise summary of the prominent news article, along with the most intuitive headline in line with the summary. When humans produce summaries of documents, they not only extract phrases and concatenate them but also produce new grammatical phrases or sentences that coincide with each other and capture the most significant information of the original article. Humans have an incredible ability to create abstractions; however, automatic summarization is a challenging problem. This paper aims to develop an end-to-end methodology that can generate brief summaries and crisp headlines that can capture the attention of readers and convey a significant amount of relevant information. In this paper, we propose a novel methodology known as SHEG, which is designed as a hybrid model. It works by integrating both extractive and abstractive mechanisms using a pipelined approach to produce a concise summary, which is then used for headline generation. Experiments were performed on publicly available datasets, viz. CNN/Daily Mail, Gigaword, and NEWSROOM. The results obtained validate our approach and demonstrate that the proposed SHEG model is effectively producing a concise summary as well as a captivating and fitting headline. © 2020, Springer-Verlag London Ltd., part of Springer Nature.