Himanshu Yadav

@manit.ac.in

Under Dr. Bholanath Roy and Dr. Vaibhav Soni, Department of Computer Science & Engineering
Maulana Azad National Institute of Technology Bhopal

Himanshu Yadav


                  Download Compact PDF  
https://researchid.co/hanny17

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Science Applications, Computer Vision and Pattern Recognition, Hardware and Architecture, Signal Processing
6

Scopus Publications

76

Scholar Citations

3

Scholar h-index

2

Scholar i10-index

Scopus Publications

  • Attention-Based Transfer Learning for Non-IID Federated Human Activity Recognition Using Inertial Sensors
    Vaibhav Soni, Mayank Gupta, Himanshu Yadav, R. K. Pateriya
    Proceedings of 2025 International Conference on Signal Processing Computation Electronics Power and Telecommunication Iconscept 2025, 2025
    Human Activity Recognition (HAR) enables health-care, fitness, and smart-environment services but must balance accuracy with privacy and on-device efficiency. Prior federated learning (FL) methods often degrade under non-IID client data and lack effective transfer of global knowledge to compact client models. We propose a hybrid FL approach that initializes client models from a centrally trained, attention-augmented BiLSTM with multi-head self-attention (MHA) and applies a lightweight feature-adaptation layer, models aggregate with FedAvg while keeping data local. We evaluate on mHealth, BandX, UCI-HAR, and WISDM using dataset-provided splits under a fixed 15-round FL protocol with evaluation based on Accuracy, Recall, Precision, and F1-score. The approach achieves 0.7479 Accuracy on UCI-HAR (+13.7 pp vs. FedAvg at 0.6107, +12.0 pp vs. FedMA at 0.6184), yields small but consistent gains on WISDM (0.6189 vs. 0.6147) and BandX (0.7048 vs. 0.7039), and accepts a privacy–accuracy trade-off on mHealth (0.9308 vs. centralized 0.9852). These results indicate that transfer-initialized, attention-based models narrow the FL–centralized gap for non-IID HAR while preserving data locality, supporting privacy-aware deployment on edge devices. Transfer learning (TL) improves client adaptation and alleviates non-IID performance degradation through the utilization of common pretrained weights.
  • A Novel Additive Attention-Based MICNN-BiLSTM Model for Fall Detection Using Wearable Inertial Sensors
    Himanshu Yadav, Divyanshu Gupta, Vaibhav Soni, Bholanath Roy
    6th IEEE International Conference on Recent Advances in Information Technology Rait 2025, 2025
    Fall detection is a critical issue in elderly care, as falls can lead to severe injuries and even fatalities. Early and accurate detection of falls is crucial to mitigate harm and reduce healthcare costs. Recent advances in deep learning have shown promise in enhancing fall detection systems by leveraging complex patterns in wearable sensor data. This paper proposes a novel model combining a multi-input convolutional neural network (MICNN), bidirectional long short-term memory (BiL-STM), and additive attention mechanism. The MICNN extracts detailed features from multiple sensors, such as accelerometers and gyroscopes, while the BiLSTM captures temporal dependencies in the data. The additive attention mechanism improves interpretability by emphasizing sensor signals that are most relevant to fall events. Moreover, the additive attentionbased MICNN-BiLSTM model demonstrates reduced validation times, enhancing its suitability for real-time applications. We evaluated the model on two benchmark datasets, KFall, and MobiAct achieving 99.30% and 98.85% accuracy, respectively, outperforming state-of-the-art methods.
  • Improving Human Activity Recognition in Smart Healthcare with Ensemble Deep Learning
    Vaibhav Soni, Himanshu Yadav, Vijay Bhaskar Semwal, Sudhakar Tripathi
    IETE Journal of Research, 2025
    Human activity recognition (HAR) leverages sensors such as accelerometers and gyroscopes to discern human physical activities, offering transformative insights for smart healthcare applications, from Parkinson's disease monitoring to diabetes management. While deep learning (DL) methods have emerged as frontrunners in HAR using wearable sensors, they often struggle with challenges stemming from long-term dependencies in sequential data and intricate feature extraction from expansive datasets. Addressing these gaps, this paper introduces a novel ensemble-based deep neural network that seamlessly integrates four distinct classifiers: CNN-LSTM, LSTM-CNN, CNN-BiLSTM, and BiLSTM-CNN. This combination enhances feature extraction and captures the detailed long-term patterns that have been challenging for traditional HAR models. By also using ensemble learning, our model becomes more consistent and reliable in its predictions. Benchmark evaluations on mHealth, UCI-HAR, and WISDM datasets validate the model's superiority, with accuracy scores of 99.91%, 98.10%, and 99.48% respectively. Further, the k-fold cross-validation technique is used to assess the performance results in terms of the mean F1-score, mean accuracy at k = 5. These compelling results underscore the model's capacity to address the inherent limitations of existing HAR systems, positioning it as a groundbreaking tool for advanced human activity recognition in smart healthcare scenarios.
  • A Hybrid Deep Learning Approach with MAOA Optimization for Enhanced Human Activity Recognition
    Vaibhav Soni, Surendra Meena, Himanshu Yadav, Akash Sinha, Vijay Bhaskar Semwal
    6th IEEE International Conference on Recent Advances in Information Technology Rait 2025, 2025
    Human Activity Recognition (HAR) plays a crucial role in improving human-computer interaction and personal health monitoring. Recent advancements in Deep Learning (DL) improve HAR performance by capturing intricate patterns in sensor data, but manual hyperparameter tuning remains a significant challenge. This process often leads to inefficiencies, including overfitting or underfitting, which negatively impact a model’s ability to generalize across diverse datasets. To address these issues, this paper introduces a DL-based approach that integrates Bidirectional Long Short-Term Memory (BiLSTM), Temporal Convolutional Network (TCN), and Convolutional Block Attention Module (CBAM) for enhanced feature extraction and sequence modeling. The Modified Arithmetic Optimization Algorithm (MAOA) is employed to automate hyperparameter tuning, improving model performance and generalization. The proposed model achieves 98.00% accuracy on the KU-HAR dataset and 99.69% on the mHEALTH dataset. Additionally, 5-fold cross-validation is applied to ensure robust evaluation, demonstrating the model’s efficiency and adaptability across varying data complexities. Our approach offers a scalable, automated solution to HAR that can be extended to other domains requiring accurate and efficient activity classification.
  • CABMNet: An adaptive two-stage deep learning network for optimized spatial and temporal analysis in fall detection
    Vaibhav Soni, Himanshu Yadav, Sadhna Bijrothiya, Vijay Bhaskar Semwal
    Biomedical Signal Processing and Control, 2024
  • A Novel Smartphone-Based Human Activity Recognition Using Deep Learning in Health care
    Vaibhav Soni, Himanshu Yadav, Vijay Bhaskar Semwal, Bholanath Roy, Dilip Kumar Choubey, Dheeresh K. Mallick
    Lecture Notes in Electrical Engineering, 2023

RECENT SCHOLAR PUBLICATIONS

  • Attention-Based Transfer Learning for Non-IID Federated Human Activity Recognition Using Inertial Sensors
    V Soni, M Gupta, H Yadav, RK Pateriya
    2025 International Conference on Signal Processing, Computation, Electronics … , 2025
    2025
  • A Hybrid Deep Learning Approach with MAOA Optimization for Enhanced Human Activity Recognition
    V Soni, S Meena, H Yadav, A Sinha, VB Semwal
    2025 6th International Conference on Recent Advances in Information … , 2025
    2025
  • A Novel Additive Attention-Based MICNN-BiLSTM Model for Fall Detection Using Wearable Inertial Sensors
    H Yadav, D Gupta, V Soni, B Roy
    2025 6th International Conference on Recent Advances in Information … , 2025
    2025
    Citations: 1
  • Improving Human Activity Recognition in Smart Healthcare with Ensemble Deep Learning
    V Soni, H Yadav, VB Semwal, S Tripathi
    IETE Journal of Research 71 (3), 894-908 , 2025
    2025
    Citations: 7
  • CABMNet: An adaptive two-stage deep learning network for optimized spatial and temporal analysis in fall detection
    V Soni, H Yadav, S Bijrothiya, VB Semwal
    Biomedical Signal Processing and Control 96, 106506 , 2024
    2024
    Citations: 23
  • A novel smartphone-based human activity recognition using deep learning in health care
    V Soni, H Yadav, VB Semwal, B Roy, DK Choubey, DK Mallick
    Machine Learning, Image Processing, Network Security and Data Sciences … , 2023
    2023
    Citations: 45

MOST CITED SCHOLAR PUBLICATIONS

  • A novel smartphone-based human activity recognition using deep learning in health care
    V Soni, H Yadav, VB Semwal, B Roy, DK Choubey, DK Mallick
    Machine Learning, Image Processing, Network Security and Data Sciences … , 2023
    2023
    Citations: 45
  • CABMNet: An adaptive two-stage deep learning network for optimized spatial and temporal analysis in fall detection
    V Soni, H Yadav, S Bijrothiya, VB Semwal
    Biomedical Signal Processing and Control 96, 106506 , 2024
    2024
    Citations: 23
  • Improving Human Activity Recognition in Smart Healthcare with Ensemble Deep Learning
    V Soni, H Yadav, VB Semwal, S Tripathi
    IETE Journal of Research 71 (3), 894-908 , 2025
    2025
    Citations: 7
  • A Novel Additive Attention-Based MICNN-BiLSTM Model for Fall Detection Using Wearable Inertial Sensors
    H Yadav, D Gupta, V Soni, B Roy
    2025 6th International Conference on Recent Advances in Information … , 2025
    2025
    Citations: 1
  • Attention-Based Transfer Learning for Non-IID Federated Human Activity Recognition Using Inertial Sensors
    V Soni, M Gupta, H Yadav, RK Pateriya
    2025 International Conference on Signal Processing, Computation, Electronics … , 2025
    2025
  • A Hybrid Deep Learning Approach with MAOA Optimization for Enhanced Human Activity Recognition
    V Soni, S Meena, H Yadav, A Sinha, VB Semwal
    2025 6th International Conference on Recent Advances in Information … , 2025
    2025