Ankit Tomar
Research Subjects
Ankit Tomar has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Santosh Kumar
Graphic Era (Deemed to be University)
4 shared publications - Kamal Kant Verma
School of Computer Science and Engineering, IILM University, India
3 shared publications Kamal Kant Verma
Lovely Professional University
2 shared publicationsPramod Kumar
Sri Ram Himalayan University
2 shared publications- Rahul Nijhawan
CSED, Thapar Institute of Engineering and Technology, Patiala, India
2 shared publications Amit Dixit
Quantum University, Roorkee, Uttarakhand
1 shared publication
Latest Publications
- Open Access
- 2 May 2025
- Published by Elsevier in Neurocomputing
- Vol. 640
EDCCN: A benchmark encoder-decoder framework for accurate crowd counting
- 2 May 2025
- Published by Elsevier in Neurocomputing
- Vol. 640
Open AccessEDCCN: A benchmark encoder-decoder framework for accurate crowd counting
- 23 December 2024
- Published by Springer Nature
- p. 169-182
People Counting via Supervised Learning-Based 2D CNN-LR Model in Complex Crowd Images
- 23 December 2024
- Published by Springer Nature
- p. 169-182
People Counting via Supervised Learning-Based 2D CNN-LR Model in Complex Crowd Images
- 23 October 2024
- Published by Springer Nature in SN Computer Science
- Vol. 5 (8) , 1-16
People Counting from Moving Camera Videos through PeopleNet Framework
- 23 October 2024
- Published by Springer Nature in SN Computer Science
- Vol. 5 (8) , 1-16
People Counting from Moving Camera Videos through PeopleNet Framework
- Open Access
- 24 March 2024
- Published by Springer Nature in Research Square
ED-DKCNN: Advancing Supervised Crowd Counting in Complex Environments
The increasing urban population has led to challenges in managing crowd dynamics, especially preventing tragic incidents like stampedes. Real-time, accurate crowd counting faces obstacles such as background clutter and perspective variations. The study accepts these challenges of supervised crowd counting by examining the effectiveness of convolutional arrangements in improving accuracy by using an encoder-decoder dynamic convolutional neural network * (ED-DKCNN). The combined segmented, edge-oriented data and texture-rich features are the input for the model capable of precise crowd counting, achievable even in complex scenarios with occlusions and dense crowds. It explores low and high-level crowd features, addresses occlusion and uneven crowd distribution, and utilizes deep mining and dense complementarity for optimized people counting without density map estimation. The proposed framework harnesses intra-and inter-depth information representation through a non-increasing-order kernel arrangement, achieving state-of-the-art accuracy in people counting compared to existing methods across various datasets. The extensive experiments over free and surveillance category datasets through multiple evaluation criteria firmly * https://link.springer.com/article/10. establish the proposed ED-DKCNN model as a state-of-the-art performer in this domain. Moreover, the proposed model significantly advances crowd-counting methodologies, offering potential applications in multi-modal data integration, real-time scenarios, privacy concerns, edge computing, cross-domain situations, and human behavior.
- 24 March 2024
- Published by Springer Nature in Research Square
Open AccessED-DKCNN: Advancing Supervised Crowd Counting in Complex Environments
The increasing urban population has led to challenges in managing crowd dynamics, especially preventing tragic incidents like stampedes. Real-time, accurate crowd counting faces obstacles such as background clutter and perspective variations. The study accepts these challenges of supervised crowd counting by examining the effectiveness of convolutional arrangements in improving accuracy by using an encoder-decoder dynamic convolutional neural network * (ED-DKCNN). The combined segmented, edge-oriented data and texture-rich features are the input for the model capable of precise crowd counting, achievable even in complex scenarios with occlusions and dense crowds. It explores low and high-level crowd features, addresses occlusion and uneven crowd distribution, and utilizes deep mining and dense complementarity for optimized people counting without density map estimation. The proposed framework harnesses intra-and inter-depth information representation through a non-increasing-order kernel arrangement, achieving state-of-the-art accuracy in people counting compared to existing methods across various datasets. The extensive experiments over free and surveillance category datasets through multiple evaluation criteria firmly * https://link.springer.com/article/10. establish the proposed ED-DKCNN model as a state-of-the-art performer in this domain. Moreover, the proposed model significantly advances crowd-counting methodologies, offering potential applications in multi-modal data integration, real-time scenarios, privacy concerns, edge computing, cross-domain situations, and human behavior.
- 25 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 537-541
HHFER: A Hybrid Framework for Human Facial Expression Recognition
Automatic recognition of human facial expressions from images or videos is considered one of the most important tasks in computer vision (CV). Some universal challenges such as dynamic illumination, different face shapes, low-resolution samples, pose variations limit the performance of facial feature recognition by conventional methods. Given these challenges, the research community has paid more attention to deep learning (DL) for automatic and accurate emotion classification. This study addresses a hybrid human facial emotion recognition (HHFER) framework, which works on a DL face-sensitive based feature extraction policy. The proposed framework extracts facial features using a CNN network and classifies them into dedicated emotions through an SVM classifier. The model has been trained on two benchmark RAF-DB and FER2013 datasets with different characteristics for seven emotion categories, for which it achieves an accuracy of 92.80%, 90.85%. Traditional testing and real-time sentiment testing on live videos prove the robustness and scalability of the proposed framework.
- 25 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 537-541
HHFER: A Hybrid Framework for Human Facial Expression Recognition
Automatic recognition of human facial expressions from images or videos is considered one of the most important tasks in computer vision (CV). Some universal challenges such as dynamic illumination, different face shapes, low-resolution samples, pose variations limit the performance of facial feature recognition by conventional methods. Given these challenges, the research community has paid more attention to deep learning (DL) for automatic and accurate emotion classification. This study addresses a hybrid human facial emotion recognition (HHFER) framework, which works on a DL face-sensitive based feature extraction policy. The proposed framework extracts facial features using a CNN network and classifies them into dedicated emotions through an SVM classifier. The model has been trained on two benchmark RAF-DB and FER2013 datasets with different characteristics for seven emotion categories, for which it achieves an accuracy of 92.80%, 90.85%. Traditional testing and real-time sentiment testing on live videos prove the robustness and scalability of the proposed framework.
- 22 April 2021
- Published by Springer Nature in Applied Intelligence
- Vol. 52 (1) , 55-70
Dynamic Kernel CNN-LR model for people counting
- 22 April 2021
- Published by Springer Nature in Applied Intelligence
- Vol. 52 (1) , 55-70
Dynamic Kernel CNN-LR model for people counting
- Open Access
- 30 December 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Engineering and Advanced Technology
- Vol. 9 (2) , 1232-1237
Improved Task Scheduling using Effective Particle Swarm Optimization in Cloud Computing Environment
A vibrant on demand service of today’s era is cloud computing where one can utilize computer resources without indirect active management by user where one can use computing resources to achieve coherence in economic scale. Since cloud computing feel like Everything as a service so there should be highly scalable and reliable mechanisms to distribute the load evenly across the VMs evenly. Innumerable cloudlet mapping policies are presented in various research articles to achieve the high performance, better QOS and minimized task execution time but maximum are conventional approaches. No unconventional realistic scheduling algorithms is available which can schedule the tasks in heterogeneous manner. Since cloudlet scheduling is crucial metrics of cloud computing that has to be heightened by combining the different parameters. This paper tried to provide effectiveness and improvement in task scheduling using nature inspired Particle Swarm optimization (PSO) strategy. A powerful nature inspired load balancing mechanism is proposed in this paper which optimized makespan and throughput in environment of varying cloudlets and virtual machines results as compared to other conventional approaches. Proposed (EPSO) algorithm is with four scheduling policies namely FCFS, Round Robin (RR) and Shortest Job First (SJF) and get near twice good throughput percentage and minimized makespan in two different environments. Author used Cloud sim toolkit and some Open Source cloud packages to simulate the results of various scheduling components. Experimental results of various components are tested and simulated on java based CloudSim toolkit framework.
- 30 December 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Engineering and Advanced Technology
- Vol. 9 (2) , 1232-1237
Open AccessImproved Task Scheduling using Effective Particle Swarm Optimization in Cloud Computing Environment
A vibrant on demand service of today’s era is cloud computing where one can utilize computer resources without indirect active management by user where one can use computing resources to achieve coherence in economic scale. Since cloud computing feel like Everything as a service so there should be highly scalable and reliable mechanisms to distribute the load evenly across the VMs evenly. Innumerable cloudlet mapping policies are presented in various research articles to achieve the high performance, better QOS and minimized task execution time but maximum are conventional approaches. No unconventional realistic scheduling algorithms is available which can schedule the tasks in heterogeneous manner. Since cloudlet scheduling is crucial metrics of cloud computing that has to be heightened by combining the different parameters. This paper tried to provide effectiveness and improvement in task scheduling using nature inspired Particle Swarm optimization (PSO) strategy. A powerful nature inspired load balancing mechanism is proposed in this paper which optimized makespan and throughput in environment of varying cloudlets and virtual machines results as compared to other conventional approaches. Proposed (EPSO) algorithm is with four scheduling policies namely FCFS, Round Robin (RR) and Shortest Job First (SJF) and get near twice good throughput percentage and minimized makespan in two different environments. Author used Cloud sim toolkit and some Open Source cloud packages to simulate the results of various scheduling components. Experimental results of various components are tested and simulated on java based CloudSim toolkit framework.
- 1 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Designing of Combinatorial Communication Networks Using Efficient Heuristic Tabu Search Algorithm
Designing a good network is a well-known NP-hard problem that involves in the selection of subset of possible links or network topology in order to minimize the network overhead and cost subdue to reliability constraint we need to design the topological framework for the fixed node network. To overcome this problem, we are using a novel algorithm based on the cloud computing techniques for designing the communication networks, considering both throughput and reliability. The proposed method is known as Tabu search algorithm. For showing the efficiency and results (compared to those obtained results from the conventional approaches)of the combinatorial TSA algorithm, a heuristic approach is applied to test with different topology network systems, i.e., genetic algorithm (GA) and Tabu search algorithm (TSA). In this article we are applying the heuristic Tabu search algorithm to maintain the reliable communication topological network in which we are comparing the performance and throughput of this algorithm with conventional Tabu search method. TSA uses the hill climbing approach for finding the optimal solution for a reliable network Various simulated results are obtained to test the problems with the various constraints. Here we have shown that the proposed approach perform better, scalable and also is superior to the conventional approaches in terms of solution quality and computational time.
- 1 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Designing of Combinatorial Communication Networks Using Efficient Heuristic Tabu Search Algorithm
Designing a good network is a well-known NP-hard problem that involves in the selection of subset of possible links or network topology in order to minimize the network overhead and cost subdue to reliability constraint we need to design the topological framework for the fixed node network. To overcome this problem, we are using a novel algorithm based on the cloud computing techniques for designing the communication networks, considering both throughput and reliability. The proposed method is known as Tabu search algorithm. For showing the efficiency and results (compared to those obtained results from the conventional approaches)of the combinatorial TSA algorithm, a heuristic approach is applied to test with different topology network systems, i.e., genetic algorithm (GA) and Tabu search algorithm (TSA). In this article we are applying the heuristic Tabu search algorithm to maintain the reliable communication topological network in which we are comparing the performance and throughput of this algorithm with conventional Tabu search method. TSA uses the hill climbing approach for finding the optimal solution for a reliable network Various simulated results are obtained to test the problems with the various constraints. Here we have shown that the proposed approach perform better, scalable and also is superior to the conventional approaches in terms of solution quality and computational time.
- Open Access
- 31 August 2016
- Published by NADIA in International Journal of Security and Its Applications
- Vol. 10 (8) , 395-406
DNS Prevention Using 64-Bit Time Synchronized Public Key Encryption to Isolate Phishing Attacks
- 31 August 2016
- Published by NADIA in International Journal of Security and Its Applications
- Vol. 10 (8) , 395-406
Open AccessDNS Prevention Using 64-Bit Time Synchronized Public Key Encryption to Isolate Phishing Attacks