Madhu Shambhu Shukla

Dr. Madhu Shambhu Shukla

Marwadi University, Rājkot, Gujarāt, India

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
31 October 2022 in Algorithms
Verified
Journal Article
Open access
PDF file
DSDebabrata Swain
Debabrata Swain
SSSantosh Kumar Satapathy
Santosh Kumar Satapathy
BABiswaranjan Acharya
Biswaranjan Acharya
MSMadhu Shambhu Shukla
Madhu Shambhu Shukla
Vassilis C. GerogiannisVassilis C. Gerogiannis
Vassilis C. Gerogiannis
DGDimitris Giakovis
Dimitris Giakovis
Activity recognition is the process of continuously monitoring a person’s activity and movement. Human posture recognition can be utilized to assemble a self-guidance practice framework that permits individuals to accurately learn and rehearse yoga postures without getting help from anyone else. With the use of deep learning algorithms, we propose an approach for the efficient detection and recognition of various yoga poses. The chosen dataset consists of 85 videos with 6 yoga postures performed by 15 participants, where the keypoints of users are extracted using the Mediapipe library. A combination of Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) has been employed for yoga pose recognition through real-time monitored videos as a deep learning model. Specifically, the CNN layer is used for the extraction of features from the keypoints and the following LSTM layer understands the occurrence of sequence of frames for predictions to be implemented. In following, the poses are classified as correct or incorrect; if a correct pose is identified, then the system will provide user the corresponding feedback through text/speech. This paper combines machine learning foundations with data structures as the synergy between these two areas can be established in the sense that machine learning techniques and especially deep learning can efficiently recognize data schemas and make them interoperable.

17 February 2025 in IEEE Transactions on Consumer Electronics
Verified
Journal Article
Closed access
AFAfifa Salsabil Fathima
Afifa Salsabil Fathima
SBSyed Muzamil Basha
Syed Muzamil Basha
Syed Thouheed AhmedSyed Thouheed Ahmed
Syed Thouheed Ahmed
SBSurbhi Bhatia
Surbhi Bhatia
FAFatima Asiri
Fatima Asiri
MSMadhu Shambhu Shukla
Madhu Shambhu Shukla
When implementing zero-trust edge computing, offloading computational tasks and data access through traditional model training and usage approaches can lead to increased latency. Since the traditional methods often involve extensive communication with a central server, creating additional network hopping stations/nodes resulting in increased latency. The challenge is bound to allocate a befitting resource at a given consumer demand. In this proposed system, a federated learning model based data offloading and consumer medical resource recommendation of IoT is discussed and validated. The user/consumer group and local training models are aligned with edge servers for data preprocessing and customization with a series of resources demand creation and coordination. The consumer resource allocating priorities are fine-grained with the proposed blockchain based priority analyzer for recommendation and allocation. The computational parameter such as resource pool, average waiting time, energy consumption and transmission trust delays are observed and validated. The proposed framework fetches consumer resources logs and synchronizes the centralized training model for effective scheduling and allocation of resources with an accuracy of 94.92% under the 5G operating spectrum. The technique has demonstrated minimal latency in offloading the data request demand and resource allocation at the cloud servers.

9 February 2024 in Information
Verified
Journal Article
Open access
PDF file
Majdi SukkarMajdi Sukkar
Majdi Sukkar
MSMadhu Shambhu Shukla
Madhu Shambhu Shukla
DKDinesh Kumar
Dinesh Kumar
Vassilis C. GerogiannisVassilis C. Gerogiannis
Vassilis C. Gerogiannis
Andreas KanavosAndreas Kanavos
Andreas Kanavos
BABiswaranjan Acharya
Biswaranjan Acharya
Effective collision risk reduction in autonomous vehicles relies on robust and straightforward pedestrian tracking. Challenges posed by occlusion and switching scenarios significantly impede the reliability of pedestrian tracking. In the current study, we strive to enhance the reliability and also the efficacy of pedestrian tracking in complex scenarios. Particularly, we introduce a new pedestrian tracking algorithm that leverages both the YOLOv8 (You Only Look Once) object detector technique and the StrongSORT algorithm, which is an advanced deep learning multi-object tracking (MOT) method. Our findings demonstrate that StrongSORT, an enhanced version of the DeepSORT MOT algorithm, substantially improves tracking accuracy through meticulous hyperparameter tuning. Overall, the experimental results reveal that the proposed algorithm is an effective and efficient method for pedestrian tracking, particularly in complex scenarios encountered in the MOT16 and MOT17 datasets. The combined use of Yolov8 and StrongSORT contributes to enhanced tracking results, emphasizing the synergistic relationship between detection and tracking modules.

28 September 2023 in Information
Verified
Journal Article
Open access
PDF file
MGManav Garg
Manav Garg
PGPranshav Ritesh Gajjar
Pranshav Ritesh Gajjar
PSPooja Shah
Pooja Shah
MSMadhu Shambhu Shukla
Madhu Shambhu Shukla
BABiswaranjan Acharya
Biswaranjan Acharya
Andreas KanavosAndreas Kanavos
Andreas Kanavos
The musical key serves as a crucial element in a piece, offering vital insights into the tonal center, harmonic structure, and chord progressions while enabling tasks such as transposition and arrangement. Moreover, accurate key estimation finds practical applications in music recommendation systems and automatic music transcription, making it relevant across academic and industrial domains. This paper presents a comprehensive comparison between standard deep learning architectures and emerging vision transformers, leveraging their success in various domains. We evaluate their performance on a specific subset of the GTZAN dataset, analyzing six different deep learning models. Our results demonstrate that DenseNet, a conventional deep learning architecture, achieves remarkable accuracy of 91.64%, outperforming vision transformers. However, we delve deeper into the analysis to shed light on the temporal characteristics of each deep learning model. Notably, the vision transformer and SWIN transformer exhibit a slight decrease in overall performance (1.82% and 2.29%, respectively), yet they demonstrate superior performance in temporal metrics compared to the DenseNet architecture. The significance of our findings lies in their contribution to the field of musical key estimation, where accurate and efficient algorithms play a pivotal role. By examining the strengths and weaknesses of deep learning architectures and vision transformers, we can gain valuable insights for practical implementations, particularly in music recommendation systems and automatic music transcription. Our research provides a foundation for future advancements and encourages further exploration in this area.

30 December 2024 in IEEE Access
Verified
Journal Article
Open access
Majdi SukkarMajdi Sukkar
Majdi Sukkar
SBSatinder Kaur Brar
Satinder Kaur Brar
MSMadhu Shambhu Shukla
Madhu Shambhu Shukla
RMRajesh Mahadeva
Rajesh Mahadeva
This paper surveys real-time object detection literature critically and analytically, focusing particularly on pedestrian detection for safe autonomous vehicles. It addresses the challenges in the domain, some of the sources of which are variations in age, gender, clothing, lighting, backgrounds, and occlusion. The paper reviews object detection algorithms after providing an overview of deep learning basics and main architectures of neural networks, followed by discussion on existing algorithms along with their strengths, weaknesses, and future research directions. There is a need for pedestrian detection datasets with further complex annotations and multi-source integration, which captures interactions between pedestrians and their surroundings. Incorporating advanced sensors, including LiDAR, infrared, and depth sensors, as the foremost means to enhance the detection capabilities in more adverse conditions, such as low-light situations and occlusion. However, architectures such as YOLO, SSD, and Faster R-CNN, which have led to current improvements in performance, still allow room for improving pedestrian detection accuracy. By filling in these insights and proposed solutions, the paper focus on the development of pedestrian detection technology, how it can be brought into a safer, reliable, real-world applicability towards the system of autonomous driving. All of these results point to continued innovation towards deep learning, multi-sensor integration, and developing datasets to achieve optimal performance levels in real world conditions for autonomous driving systems.

18 August 2022 in Computational Intelligence and Neuroscience
Verified
Journal Article
Open access
PDF file
KZKirtirajsinh Zala
Kirtirajsinh Zala
HTHIREN KUMAR THAKKAR
HIREN KUMAR THAKKAR
SBSatinder Kaur Brar
Satinder Kaur Brar
NDNeel H. Dholakia
Neel H. Dholakia
Ketan KotechaKetan Kotecha
Ketan Kotecha
MSMadhu Shambhu Shukla
Madhu Shambhu Shukla
Traditional healthcare services have changed into modern ones in which doctors can diagnose patients from a distance. All stakeholders, including patients, ward boy, life insurance agents, physicians, and others, have easy access to patients medical records due to cloud computing. The clouds services are very cost-effective and scalable, and provide various mobile access options for a patients electronic health records (EHRs). EHR privacy and security are critical concerns despite the many benefits of the cloud. Patient health information is extremely sensitive and important, and sending it over an unencrypted wireless media raises a number of security hazards. This study suggests an innovative and secure access system for cloud-based electronic healthcare services storing patient health records in a third-party cloud service provider. The research considers the remote healthcare requirements for maintaining patient information integrity, confidentiality, and security. There will be fewer attacks on e-healthcare records now that stakeholders will have a safe interface and data on the cloud will not be accessible to them. End-to-end encryption is ensured by using multiple keys generated by the key conclusion function (KCF), and access to cloud services is granted based on a persons identity and the relationship between the parties involved, which protects their personal information that is the methodology used in the proposed scheme. The proposed scheme is best suited for cloud-based e-healthcare services because of its simplicity and robustness. Using different Amazon EC2 hosting options, we examine how well our cloud-based web application service works when the number of requests linearly increases. The performance of our web application service that runs in the cloud is based on how many requests it can handle per second while keeping its response time constant. The proposed secure access scheme for cloud-based web applications was compared to the Ethereum blockchain platform, which uses internet of things (IoT) devices in terms of execution time, throughput, and latency.