About

I received my B.Sc. (Engg.) in Electrical Engineering in 2009 from BIT Sindri Dhanbad, M.Tech. in Instrumentation in 2011 from Indian Institute of Technology, Kharagpur, and PhD in Electrical Engineering from Indian Institute of Technology, Kanpur in 2024 with research focus on algorithmic development for efficient AI models and their applications to modern industrial intelligent systems. I am currently working as senior project executive at department of electrical engineering, IIT Kanpur, where, I am working on novel lightweight deep learning models for 6G standardization. Previously I have worked at SMSS Lab, ITT Kanpur where, I have developed new AI benchmarks for inverse design of mechanical metastructures, POC-based cross-spectral vision transformer for face biometrics, and multi-dynamics regression model for robotic arm. My research interests are inclined towards developing AI algorithms for real-time, interdisciplinary autonomous systems.

I have a strong background in machine learning, deep learning, and generative models applied to interdisciplinary engineering problems, especially intelligent diagnostic systems, computer vision, Image-based analysis, and inverse design of smart structures and systems.

Efficient AI

Efficient Training of DL Model

Guided Sampling

PG-Gradient-based Guided sampling for Network Architecture Search

Medical Vision

Computer Vision for Medical Imaging Classification

Engineering Design

AI in Engineering Design (Inverse Design of Metastructure)

POC-based Cross-Spectral Attention

POC-based CSA: A new benchmark of phase only based feature extrction

Research Highlights

Deep learning algorithms have shown outstanding performance for system identification, fault diagnosis, classification, feature extraction, etc. However, the performance of most of the deep learning methodologies is greatly affected by the selection of suitable optimization techniques, model architecture, and model hyper-parameters. My Ph.D. research has attempted to develop efficient methods of model selection with a novel fitness evaluation strategy for improved performance in real-time scenarios. Fast model evaluation is one of the key requirements of model selection algorithms. Therefore, methods for fast model evaluation strategies are developed to reduce training time and the requirements of computational resources.

Ph.D. Research Highlights

  • Developed an improved and faster parameter optimization in parallel computational framework for "Aerodynamic Modelling of ATTAS Aircraft using Mamdani Fuzzy Inference Network".
  • Developed a Quick Learning Mechanism with Cross-Domain Adaptation for Intelligent Fault Diagnosis — enabling fast model evaluation required for model selection.
  • Proposed EvoN2N: Knowledge transfer based Evolutionary Deep Neural Network for architecture optimization in Intelligent Fault Diagnosis.
  • Proposed GS-EvoN2N: Guided Sampling-based Evolutionary Deep Neural Network; an extention of EvoN2N with policy gradient-based controller for guided sampling of neural architecture.
  • Introduced A Novel Vision Transformer with Residual in Self-attention for Biomedical Image Classification.

Research highlights while partly associated with Avermass GmbH, Germany (Remote)

  • Face Forgery Detection: Cascaded Twin models (ViT and YOLOv7) achieved 95% accuracy in forgery detection.
  • Image Data Synthesis: Implemented DCGAN and CycleGAN for generating synthetic face data and image-to-image translation.
  • End2End Image Forgery Detection: Combined ViT and YOLOv7 for fake news detection in forged images.

Postdoctoral Research at SMSS Lab, IIT Kanpur

  • AI-driven Inverse Design of Band-Tunable Mechanical Metastructures for Tailored Vibration Mitigation. arXiv:2412.12122
  • Cross-Spectral Vision Transformer for Biometric Authentication using Subcutaneous Vein and Periocular Pattern. arXiv:2412.19160
  • Multi-Dynamics Attention-Based Regression Model for Deflection Angle Prediction in SMA-Powered Metamaterial-Based Elbow Joints.

Future Research Plans

  • Short-Term: 1. Efficient and sustanable AI for engineering application. 2. AI-driven bioinformatics for diagnostics and soft-robotics for adaptive medical applications.
  • Long-Term: Autonomous AI-enabled soft robotic systems for surgical and rehabilitation applications using bio-inspired reinforcement learning models.

Publications

1. Journal Publications

  1. Arun K. Sharma and Nishchal K. Verma. “Guided Sampling-based Evolutionary Deep Neural Network for Intelligent Fault Diagnosis.” Engineering Applications of Artificial Intelligence, Volume 128, 2024. DOI | Download PDF
  2. Arun K. Sharma and Nishchal K. Verma. “Quick Learning Mechanism with Cross-Domain Adaptation for Intelligent Fault Diagnosis.” IEEE Transactions on Artificial Intelligence, Volume 3, Issue 3, June 2022. DOI | Download PDF
  3. Arun K. Sharma, Dhan Jeet Singh, V. Singh, N. K. Verma. “Aerodynamic Modelling of ATTAS Aircraft using Mamdani Fuzzy Inference Network.” IEEE Transactions on Aerospace and Electronic Systems, Volume 56, Issue 5, October 2020. DOI | Download PDF
  4. A. Dwivedi, G. Ray, Arun K. Sharma. “Genetic Algorithm Based Decentralized PI Type Controller: Load Frequency Control.” Journal of The Institution of Engineers (India): Series B, May 2015. DOI | Download PDF
  5. Arun K. Sharma, G. Ray. “Robust Controller with State-Parameter Estimation Algorithm for Uncertain Networked Control System.” IET Control Theory & Applications, Volume 6, Issue 18, December 2012. DOI | Download PDF

2. Conference Publications

  1. Arun K. Sharma, Dhanjeet Singh, and Nishchal K. Verma. “Data Driven Aerodynamic Modeling Using Mamdani Fuzzy Inference Systems.” 2018 International Conference on Sensing, Diagnostics, Prognostics, and Control, Xi’an, China.
  2. Arun K. Sharma, Vikas Singh, Nishchal K. Verma, Jie Liu. “Condition Based Monitoring of Machine using Mamdani Fuzzy Network.” 2018 Prognostics and System Health Management Conference, PHM-Chongqing, China.
  3. Arun K. Sharma, S. Pandey. “Automatic Generation Control of Interconnected Power System using Internal & Variable System Controller.” 2nd National Conference on Advancement of Electronics & Communication Technology and Engineering, Greater Noida, 2012.
  4. Arun K. Sharma. “Adaptive Algorithm for State and Parameter Estimation in a system with Parametric Uncertainties.” National Conference on Advances in Electrical Power and Energy Systems, AKGEC, Ghaziabad, 2013.

3. Preprints

  1. Arun K. Sharma, Shubhobrata Bhattacharya, Motahar Reza, Bishakh Bhattacharya. “Cross-Spectral Vision Transformer for Biometric Authentication using Forehead Subcutaneous Vein Pattern and Periocular Pattern.” arXiv preprint, arXiv:2412.19160.
  2. T. Gupta, Arun K. Sharma, A. Dwivedi, V. Gupta, S. Sahana, S. Pathak, A. Awasthi, B. Bhattacharya. “AI-driven Inverse Design of Band-Tunable Mechanical Metastructures for Tailored Vibration Mitigation.” arXiv preprint, arXiv:2412.12122.
  3. Arun K. Sharma, Nishchal K. Verma. “A Novel Vision Transformer with Residual in Self-attention for Biomedical Image Classification.” arXiv preprint, arXiv:2306.01594.
  4. Arun K. Sharma, Nishchal K. Verma. “Knowledge Transfer based Evolutionary Deep Neural Network for Intelligent Fault Diagnosis.” arXiv preprint, arXiv:2109.13479.

Experiences

  • Postdoctoral Researcher, Intelligent Wireless Lab, Department of Electrical Engineering, Indian Institute of Technology, Kanpur (May 2025–Present, Supervisor: Prof. Rohit Budhiraja)
    Working on development of novel AI for 6G standardization.
  • Postdoctoral Researcher, SMSS Lab, Department of Mechanical Engineering, Indian Institute of Technology, Kanpur (Sept. 2024–April 2025, Supervisor: Prof. Bishakh Bhattacharya)
    WOrked on development of new AI algorithms for engineering design, including inverse design and intelligent actuators.
  • Lead AI Scientist (Remote, Partly), Avermass GmbH, German startup (June 2022–May 2023)
    Led ML/DL research team for various AI-based projects: forgery detection, synthetic data generation, and Document AI.
  • Assistant Professor, IIMT College of Engineering, Greater Noida (July 2011–July 2013)
    Teaching and Research in Electrical and Instrumentation Engineering.
  • Assistant Professor, AKGEC, Ghaziabad (July 2013–Nov. 2017)
    Teaching and Research in Electrical and Electronic Engineering. Led student projects on AI, ML applications, and lab development.
  • Teaching Assistant, IIT Kanpur (During PhD)
    Involved in teaching Control Systems, Electrical Machines, AI/ML, and related labs.

Total: Inds. startup = 01 year, Teaching = 6.5 year, Post-Doc = 10 months

Subjects Taught

  • Control Systems & Applications
  • Electrical Machines & Power Systems
  • Measurement & Instrumentation
  • Artificial Intelligence and Machine Learning
  • Sensor and Transducer Systems
  • LabVIEW-based Instrumentation

Timeline

May 2025–Present

Post-Doctoral Researcher at Intelligent Wireless Lab, EE, Indian Institute of Technology, Kanpur, UP, India.

Responsibility: Research on New AI for 6G standardization under Prof. Rohit Budhiraja

Sept. 2024–April 2025

Post-Doctoral Researcher at SMSS Lab, Indian Institute of Technology, Kanpur, UP, India.

Responsibility: Research on New AI for Engineering applications (1. Inverse Design of Metastructure, 2. POC-based new attention for face biometrics, 3. Multimodal deep learning model for bionic arm actuation) under Prof. Bishakh Bhattacharya

Jan. 2018–Aug. 2024

Ph.D. in Electrical Engineering, Indian Institute of Technology, Kanpur, UP, India.

Thesis: "Deep Learning Architecture Optimization and Application to Classification Problems"

July 2013-Nov. 2017

Assistant Professor, AKGEC, Ghaziabad, India.

Responsibility: Teaching and Reasearch

July 2011-July 2013

Assistant Professor, IIMT College of Engineering, Greater Noida, India.

Responsibility: Teaching and Reasearch

July 2009-June 2011

Master of Tehcnology in Electrical Engineering,

Indian Institute of Technology, Kharagpur, India.

Thesis: "Estimator Based Robust Stabilization of Uncertain Networked Control System"

July 2005-June 2009

Bachelor of Technology, BIT Sindri, Dhanbad, Jharkhand, India.

University: Vinoba Bhave University, Hazaribagh

Contact

    Address: Dr. Arun Kumar Sharma, 113, ACES, EE, IIT Kanpur, India

    Email: arunshr.iit@gmail.com