(SoE) School of Engineering

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.

Shiv Nadar University

We propose a novel video sampling scheme for human action recognition in videos, using Gaussian Weighing Function. Traditionally in deep learning-based human activity recognition approaches, either a few random frames or every $k^{th}$ frame of the video is considered for training the 3D CNN, where $k$ is a small positive integer, like $4$, $5$, or $6$. This kind of sampling reduces the volume of the input data, which speeds-up the training network and also avoids overfitting to some extent, thus enhancing the performance of the 3D CNN model. In the proposed video sampling technique, consecutive $k$ frames of a video are aggregated into a single frame by computing a Gaussian-weighted summation of the $k$ frames. The resulting frame preserves the information in a better way than the conventional approaches and experimentally shown to perform better. In this paper, a 3-Dimensional deep CNN is proposed to extract the spatio-temporal features and follows Long Short-Term Memory (LSTM) to recognize human actions. The proposed 3D CNN architecture is capable of handling the videos where the camera is placed at a distance from the performer. Experiments are performed with KTH, WEIZMANN, and CASIA-B Human Activity and Gait datasets, whereby it is shown to outperform state-of-the-art deep learning based techniques. We achieve 95.78%, 95.27%, and 95.27% over the KTH, WEIZMANN, and CASIA-B human action and gait recognition datasets, respectively. &nbsp;&nbsp;

World University root node

Computer Science and Engineering

Multimedia Tools and Applications

An Information-rich Sampling Technique over Spatio-Temporal CNN for Classification of Human Actions in Videos

Mechanical Engineering

Procedia CIRP

A Dynamic System to Predict an Assembly Line Worker’s Comfortable Work-Duration Time by Using the Machine Learning Technique

Environmental Science and Pollution Research

Review on performance assessment of solar stills using computational fluid dynamics (CFD)

Fulltext

Scientific Reports

Author Correction: Novel hybrid method to additively manufacture denser graphite structures using Binder Jetting (Scientific Reports, (2021), 11, 1, (2438), 10.1038/s41598-021-81861-w)

Cognitive activity prediction (CAP) from electroencephalogram (EEG) signals is progressively utilized in the field of brain-computer interface (BCI) and mental health management. Various machine and deep learning methods have been proposed recently for CAP. However, since Internet-of-Things-based real-time BCI systems demand low latency, power, and portability, these methods need to be deployable on resource-constrained edge devices. Towards this aspect, we propose a real-time implementation of a lightweight 1-D convolutional neural network on an Arduino Due microcontroller for CAP from EEG signals. The performance evaluation on two public datasets and one real-time recorded dataset indicates that the proposed work achieves subject-independent prediction accuracies of 99.30%, 82.50%, and 99.02% in these datasets. Furthermore, the prediction of real-time recorded EEG signals is accurate for majority of the subjects. The proposed work outperforms the existing techniques and achieves low power consumption of $\text{0.63}\,\text{W}$ in real-time on-device implementation with an average latency of $\text{455.12}\,\text{ms}$ in model deployment, test output prediction, and activity-based transmission. © 2017 IEEE.

Electrical Engineering

IEEE Sensors Letters

On-Device Implementation for Deep-Learning-Based Cognitive Activity Prediction

Food waste generation is a worldwide phenomenon and disposing off it in an environmentally benign way has been a challenge. Thermochemical processes have the potential for not only processing mixed food waste effectively from an environmental point of view but also producing bioenergy in all three forms: solid (biochar), liquid (bio-oil) and gas (syngas). In this study, two thermochemical processes – fast pyrolysis and steam gasification – aimed for producing syngas as main product were carried out at three different temperatures: 600°C, 700°C and 800°C, and resulting syngas was characterised and compared for syngas yield, syngas composition, hydrogen yield and high heating value (HHV). The steam flow rate (SFR) was maintained at 0.625 mL min−1 for all gasification experiments. The syngas yield obtained from steam gasification was higher (1.2 m3 kg−1) than the syngas yield from fast pyrolysis (0.81 m3 kg−1). In addition, the hydrogen fraction was much higher in syngas from steam gasification (63.58%) than that from fast pyrolysis (45.03%). Furthermore, carbon conversion efficiency (CCE) and apparent thermal efficiency (ATE) were determined to compare the performance of these two processes. CCE was higher (63.6%) for steam gasification than that for pyrolysis (52.3%) which suggested that steam gasification was much more effective than fast pyrolysis to produce syngas of higher quality. © The Author(s) 2022.

Chemical Engineering

Waste Management and Research

Syngas production from fast pyrolysis and steam gasification of mixed food waste

This study introduced an integrated two-stage process (ITP) for treating the mixed food waste (MFW) and producing syngas with higher hydrogen fraction where in-situ steam gasification was conducted after fast pyrolysis at the same temperature. Characteristics of syngas (syngas yield, composition, H2 yield, and high heating value (HHV)) from the proposed approach was first compared with that from conventional overlapping process (COP) of steam gasification process at the same temperature of 700°C and then with that from ITP at a higher temperature of 850°C. Syngas yield from ITP at 700°C (0.91 m3/kg) was almost the same as the syngas yield from COP at 700°C (0.85 m3/kg); however, the hydrogen and CO fraction were much higher (H2: gasification part~ 63%, average ~ 57.8%; CO: gasification part ~ 25%, average ~ 24%) than that from COP at 700°C (H2: 52%, CO: 19%). The hydrogen fraction in syngas further increased to 68% (from the gasification part) on increasing the ITP temperature to 850. Moreover, the syngas yield increased significantly (1.51 m3/kg) at the higher temperature of 850°. Finally, change in HHV values and corresponding changes in cold gas efficiency (CGE) were also presented to evaluate the process performance. © 2022 Taylor & Francis Group, LLC.

Energy Sources, Part A: Recovery, Utilization and Environmental Effects

Hydrogen rich syngas production from food waste via an integrated two-stage process of in-situ steam gasification after fast pyrolysis

Road accidents caused by several factors are responsible for the deaths and injuries of people. The majority of deaths occur within first few hours of an accident. If in case of an accident the victim becomes incapacitated, and there is no onlooker to help, then an accurate automated accident reporting system can significantly reduce the number of such deaths. This paper reports an Android smartphone-based end-to-end Internet of Things (IoT) system that can transmit accident information to emergency services and affected families once a vehicle accident is detected. Along with the detection, the classification of vehicle accidents can be very helpful in identifying suitable life-saving medical-aid and appropriate rescue operation equipment. The main objective of this research work is to develop a machine learning (ML) model that can detect as well as classify vehicle accidents accurately into eight categories. To enhance the classification efficacy of the system, a multi-sensor fusion framework has been proposed that incorporates various sensor-fusion techniques at different levels of its implementation along with several preprocessing methods such as 10-ms moving maximum, complementary filters, and a 2-sec sliding window. The framework uses Logistic Regression (LR) based stacked generalization approach to combine the decisions of three ML classifiers, which are based on Decision Tree (DT), Naïve Bayes (NB), and Random Forest (RF) methods. The LR-based stacking classifier is found to be very accurate with an F1-score of 0.95, whose performance is significantly better than the performance of all individual base-classifiers. © 2022 Elsevier B.V.

Future Generation Computer Systems

Vehicle accident sub-classification modeling using stacked generalization: A multisensor fusion approach

The increased emphasis on vehicle and occupant safety in automotive accidents has necessitated research on vehicle crashworthiness during the design phase. The design of the bumper plays a significant role in mitigating pedestrian injuries and the risk of life during accidents. The present study is focused on the design and development of an economical external multilayered bumper that can be retrofitted to any car’s existing bumper fascia while improving the safety of pedestrians without compromising the safety of engine components. The proposed multilayered bumper consists of spring and damper assembly along with a layer of high-density polyurethane foam. Finite element analysis has been performed to compute the total deformation, equivalent Von-Misses stress, spring–damping coefficients, and energy absorption using the proposed bumper. An empirical relation amongst considered design parameters is formulated, which can be used to further develop the proposed modified bumper in any feasible ways. A comparative study has also been carried out between the crashworthiness of the conventional bumper and the developed bumper at various speeds. It is observed that a significant reduction of approximately 40% in the impact is observed with the proposed bumper compared to the conventional bumper. © 2021 Informa UK Limited, trading as Taylor & Francis Group.

International Journal of Crashworthiness

Revamping the vehicle protection system using next-generation car bumper to mitigate pedestrian fatality

Poorly constructed roads, as well as roads that are damaged with cracks or potholes, are some of the leading causes of road accidents, particularly in developing nations like India. Accurate assessment of road conditions and prompt transmission of such situations may allow interested agencies to take corrective action, reducing the number of road accidents and fatalities. With the rapid advancements in sensor technologies, data aggregation, and fusion algorithms, it is now possible to identify such road conditions and report them for prompt action by the concerned authorities. Using mobile sensing, this study aims to collect and monitor the road surface conditions in the city of Dehradun, Uttarakhand, India. In order to identify different road surfaces such as smooth roads, speed hump, and rumble strips, the collected data has been analyzed using the Artificial Neural Network technique. The proposed model has an accuracy of 97%, demonstrating the algorithm's effectiveness in detecting road surface conditions. © 2022 IEEE.