Tej Singh

@mitsgwalior.in

Assistant Professor, Department of Information Technology
Madhav Institute of Science and Technology, Gwalior

Tej Singh


                     Download Compact PDF  
https://researchid.co/ttomar07
Tej Singh received B.Tech. Degree from Madan Mohan Malaviya University of Technology, Gorakhpur, (Formerly known as MMM ENGG. COLLEGE), India, in 2010, an M.E degree from the Thapar Institute of Engineering and Technology, Patiala, India, and a Ph.D. from Delhi Technological University (DTU) Delhi, India, in 2014 and 2020 respectively. He is currently working as an Assistant Professor in the Department of Information Technology, MITS Gwalior, M.P, India. His current research interests include image processing, pattern analysis, machine learning, deep learning human, artificial intelligence action, and activity recognition. He is a reviewer of various IEEE, IET, and Elsevier Journals. --Received a "Premiere Research Award" in 2021 --Received a "Commendable Research Award" in 2020 --Travel Grant to attend IEEE BigMM at NUSS, Singapore, Sept 2019.

EDUCATION

Tej Singh received B.Tech. Degree from Madan Mohan Malaviya University of Technology, Gorakhpur, (Formerly known as MMM ENGG. COLLEGE), India, in 2010, an M.E degree from the Thapar Institute of Engineering and Technology, Patiala, India, and a Ph.D. from Delhi Technological University (DTU) Delhi, India, in 2014 and 2020 respectively.

RESEARCH INTERESTS

Computer Vision, Machine Learning, Deep Learning
11

Scopus Publications

356

Scholar Citations

6

Scholar h-index

6

Scholar i10-index

Scopus Publications

  • Deep Learning-Based Human Activity Recognition Using Smartphone Sensor
    Ishant Singh, Vaibhav Anuragi, Tanisha Jain, Tej Singh
    2025 IEEE 4th World Conference on Applied Intelligence and Computing Aic 2025, 2025
    Human Activity Recognition (HAR) employing body-worn sensors particularly on the smartphone is attracting interest in various fields including, health, physical training as well as behavioral diagnosis. This work uses DeepConvLSTM model for the classification of human activities from the data collected through smart phone sensors including the accelerometer, gyroscope, and gravity sensors. In the model development, the UCI HAR dataset which involves six activities namely sitting, standing, walking, walking upstairs, walking downstairs and laying were used. To improve algorithm computational speed while maintaining data consistency. The proposed model deals with time-based sensor data in windows which incorporates both short term and long-term temporal correlations with the sensor data and achieves an accuracy of nearly 96 percent. This demonstrates the advantage of the presented model and the possibility of its future modification for even more accurate tracking of human activity compared to standard HAR systems
  • Safety Measures in Smart Car Using IoT
    Piyush Panthi, Prateek Singh Gurjar, Tej Singh, Ankit Tiwari
    Lecture Notes in Electrical Engineering, 2025
  • Speech Emotion Recognition using CNN-TRANSFORMER Architecture
    Sarthak Mangalmurti, Ojshav Saxena, Tej Singh
    2024 IEEE International Conference on Interdisciplinary Approaches in Technology and Management for Social Innovation Iatmsi 2024, 2024
    In the past few years, recognition of emotions from speech has become increasingly popular, due to its wide-ranging potential applications across various fields such as healthcare, and entertainment. The capability to identify and understand human emotions through speech is an intriguing and important research area, offering significant implications for a variety of practical uses. Sentiment analysis has garnered considerable attention due to its potential to enhance human-computer interaction, create more empathetic AI systems, and improve the quality of healthcare and entertainment experiences. In the contemporary era, Speech Emotion Recognition (SER) is increasingly vital, with its applications spanning across human-computer interactions for more empathetic AI, early detection of mental health issues, personalized content and education, market research insights, content creation, and improved accessibility. SER is paramount in addressing the evolving needs of technology, healthcare, education, and entertainment, contributing to a more inclusive, emotionally intelligent, and interconnected digital world.
  • Circulate Matrix and Compression Sensing Based Multi-Level Image Encryption
    Ranjeet Kumar Singh, Ganesh Gupta, Tej Singh, Kalka Dubey, Anjula Mehto
    Traitement Du Signal, 2022
    Digital data security is a broad research area in the field of science and technology. A lot of research was focused on information security-based mechanism for secure communication. This paper presents a novel image encryption as well as compression based on measurement matrix, pixel exchange and logistic cat map, which includes the permutation, compression, and diffusion processes. Initially the image is divided into four equal sizes of blocks and then each block is transformed into horizontal and vertical low and high frequency band. Then a random matrix multiplication function is applied to achieve an encrypted and scrambling frequency component and apply inverse DWT procedure to get first level of scrambled blocks, and further we apply the second level of security mechanism. Here each adjacent block pixel is exchanged by using the random matrices. For providing the high level of compression we design measurement matrices in compressive sensing by utilizing the circulate matrices and controlling the original column vectors of the circulate matrices with Arnold cat map. With the help of measurement matrix again the blocks are encrypted. Experimental results and performance analyses validate the good compression performance and high security of the given algorithm.
  • A deep multimodal network based on bottleneck layer features fusion for action recognition
    Tej Singh, Dinesh Kumar Vishwakarma
    Multimedia Tools and Applications, 2021
  • A deeply coupled ConvNet for human activity recognition using dynamic and RGB images
    Tej Singh, Dinesh Kumar Vishwakarma
    Neural Computing and Applications, 2021
  • Deep learning framework for single and dyadic human activity recognition
    Tej Singh, Shivam Rustagi, Aakash Garg, Dinesh Kumar Vishwakarma
    Proceedings 2019 IEEE 5th International Conference on Multimedia Big Data Bigmm 2019, 2019
    Recently, human activity recognition in videos attracts much attention in the computer vision community because of its broad real-life applications. In this context, we introduced a robust two-stream deep learning model with less complexity which utilized only the raw RGB color sequences and their dynamic motion images (DMIs) to recognize complex human activities. The RGB frames are trained using a pre-trained Inception-v3 network and CNN-LSTM with end-to-end training and for dynamic image stream, we fine-tuned the last few layers of the pre-trained model. Through our two-stream network, the features extracted from both, are max fused to increase the classification accuracy. The proposed approach has been evaluated over single-person activity dataset MIVIA Action as well as dyadic SBU Interaction dataset. Our model obtained significant performance improvement over existing similar approaches.
  • Video benchmarks of human action datasets: a review
    Tej Singh, Dinesh Kumar Vishwakarma
    Artificial Intelligence Review, 2019
  • A visual cognizance based multi-resolution descriptor for human action recognition using key pose
    Dinesh Kumar Vishwakarma, Tej Singh
    AEU International Journal of Electronics and Communications, 2019
  • Human activity recognition in video benchmarks: A survey
    Tej Singh, Dinesh Kumar Vishwakarma
    Lecture Notes in Electrical Engineering, 2019
  • A Hybrid Neuro-Wavelet Based Pre-Processing Technique for Data Representation
    Tej Singh, D.K Vishwakarma
    2017 IEEE International Conference on Computational Intelligence and Computing Research Iccic 2017, 2018

RECENT SCHOLAR PUBLICATIONS

  • Safety Measures in Smart Car Using IoT
    A Panthi, P., Gurjar, P.S., Singh, T., Tiwari
    Proceedings of the International Conference on Sensors and Microsystems … , 2025
    2025
  • Deep Learning-Based Human Activity Recognition Using Smartphone Sensor
    I Singh, V Anuragi, T Jain, T Singh
    2025 IEEE 4th World Conference on Applied Intelligence and Computing (AIC) , 2025
    2025
  • A CNN Framework for COVID Identification Using Radiographic Images
    D Jain, A Jain, T Singh
    International Conference on Generative Artificial Intelligence, Cryptography … , 2024
    2024
  • Speech Emotion Recognition using CNN-TRANSFORMER Architecture
    S Mangalmurti, O Saxena, T Singh
    2024 IEEE International Conference on Interdisciplinary Approaches in … , 2024
    2024
    Citations: 3
  • Circulate Matrix and Compression Sensing Based Multi-Level Image Encryption
    RK Singh, G Gupta, T Singh, K Dubey, A Mehto
    Traitement du Signal 39 (3), 853-862 , 2022
    2022
    Citations: 1
  • A deep multimodal network based on bottleneck layer features fusion for action recognition
    T Singh, DK Vishwakarma
    Multimedia Tools and Applications , 2021
    2021
    Citations: 17
  • A deeply coupled ConvNet for human activity recognition using dynamic and RGB images
    T Singh, DK Vishwakarma
    Neural Computing and Applications 33 (1), 469-485 , 2021
    2021
    Citations: 122
  • Deep Learning Framework for Single and Dyadic Human Activity Recognition
    T Singh, S Rustagi, A Garg, DK Vishwakarma
    2019 IEEE Fifth International Conference on Multimedia Big Data (BigMM), 237-241 , 2019
    2019
    Citations: 3
  • Video benchmarks of human action datasets: a review
    T Singh, DK Vishwakarma
    Artificial Intelligence Review 52 (2), 1107-1154 , 2019
    2019
    Citations: 104
  • A visual cognizance based multi-resolution descriptor for human action recognition using key pose
    DK Vishwakarma, T Singh
    AEU-International Journal of Electronics and Communications 107, 157-169 , 2019
    2019
    Citations: 33
  • A hybrid framework for action recognition in low-quality video sequences
    T Singh, DK Vishwakarma
    arXiv preprint arXiv:1903.04090 , 2019
    2019
    Citations: 12
  • Human activity recognition in video benchmarks: A survey
    T Singh, DK Vishwakarma
    Advances in Signal Processing and Communication: Select Proceedings of ICSC … , 2018
    2018
    Citations: 60
  • A hybrid neuro-wavelet based pre-processing technique for data representation
    T Singh, DK Vishwakarma
    2017 IEEE International Conference on Computational Intelligence and … , 2017
    2017
    Citations: 1
  • Application of Data Pre-Processing Techniques for Supervised Classification
    T Singh, RG Kumar
    2014

MOST CITED SCHOLAR PUBLICATIONS

  • A deeply coupled ConvNet for human activity recognition using dynamic and RGB images
    T Singh, DK Vishwakarma
    Neural Computing and Applications 33 (1), 469-485 , 2021
    2021
    Citations: 122
  • Video benchmarks of human action datasets: a review
    T Singh, DK Vishwakarma
    Artificial Intelligence Review 52 (2), 1107-1154 , 2019
    2019
    Citations: 104
  • Human activity recognition in video benchmarks: A survey
    T Singh, DK Vishwakarma
    Advances in Signal Processing and Communication: Select Proceedings of ICSC … , 2018
    2018
    Citations: 60
  • A visual cognizance based multi-resolution descriptor for human action recognition using key pose
    DK Vishwakarma, T Singh
    AEU-International Journal of Electronics and Communications 107, 157-169 , 2019
    2019
    Citations: 33
  • A deep multimodal network based on bottleneck layer features fusion for action recognition
    T Singh, DK Vishwakarma
    Multimedia Tools and Applications , 2021
    2021
    Citations: 17
  • A hybrid framework for action recognition in low-quality video sequences
    T Singh, DK Vishwakarma
    arXiv preprint arXiv:1903.04090 , 2019
    2019
    Citations: 12
  • Speech Emotion Recognition using CNN-TRANSFORMER Architecture
    S Mangalmurti, O Saxena, T Singh
    2024 IEEE International Conference on Interdisciplinary Approaches in … , 2024
    2024
    Citations: 3
  • Deep Learning Framework for Single and Dyadic Human Activity Recognition
    T Singh, S Rustagi, A Garg, DK Vishwakarma
    2019 IEEE Fifth International Conference on Multimedia Big Data (BigMM), 237-241 , 2019
    2019
    Citations: 3
  • Circulate Matrix and Compression Sensing Based Multi-Level Image Encryption
    RK Singh, G Gupta, T Singh, K Dubey, A Mehto
    Traitement du Signal 39 (3), 853-862 , 2022
    2022
    Citations: 1
  • A hybrid neuro-wavelet based pre-processing technique for data representation
    T Singh, DK Vishwakarma
    2017 IEEE International Conference on Computational Intelligence and … , 2017
    2017
    Citations: 1
  • Safety Measures in Smart Car Using IoT
    A Panthi, P., Gurjar, P.S., Singh, T., Tiwari
    Proceedings of the International Conference on Sensors and Microsystems … , 2025
    2025
  • Deep Learning-Based Human Activity Recognition Using Smartphone Sensor
    I Singh, V Anuragi, T Jain, T Singh
    2025 IEEE 4th World Conference on Applied Intelligence and Computing (AIC) , 2025
    2025
  • A CNN Framework for COVID Identification Using Radiographic Images
    D Jain, A Jain, T Singh
    International Conference on Generative Artificial Intelligence, Cryptography … , 2024
    2024
  • Application of Data Pre-Processing Techniques for Supervised Classification
    T Singh, RG Kumar
    2014