Tutorials

•       ➤ Click Here To know more.

Abstract: Recent advances in the field of AI in general and Deep Learning in particular have mesmerized the world due to their state-of-the-art performance in almost every aspect of science and engineering. Considering the myriad applications of AI , research communities across the globe are working towards producing better and better deep neural network architectures which are surpassing the performance of the pre-existing techniques in terms of accuracy, prediction speed, or computational efficiency. Be it a process industry, healthcare industry, genomics, medicine, engineering, robotics, or agriculture, everywhere deep learning has got its applications. But most of these deep learning models are black-box and thus are difficult to get interpreted by practicing engineers. Moreover, AI practitioners and researchers are more worried about the explainability, reliability, dependability, and trustworthiness of these well- tested, well-validated black-box AI models which are susceptible to getting fooled easily just with a naive perturbation in the sample data or with little change in model's parameters. In this tutorial, a broad overview of different practical AI problems would be discussed which make the deep learning models vulnerable in the real-world applications. The tutorial will also give an idea of available defense mechanisms which may help to minimize the risk of befooling the well- trained deep learning models in a real scenario and also can help to make them reliable and robust enough to deal with different adversaries during the deployment phase.

Short Bio-data: Dr. Rajesh P. Barnwal is serving as a Principal Scientist of CSIR-India and currently leading the AI & IoT Lab at CSIR-Central Mechanical Engineering Research Institute, Durgapur. He has more than 16 years of professional and R&D experience in the field of Computer Science and allied disciplines. He has earned his M.Tech. and Ph.D. degrees from the Indian Institute of Technology (IIT), Kharagpur, India. He was awarded the Senior Member award by the Association of Computing Machinery (ACM), USA in the year 2013 and was elevated to Senior Member grade of IEEE in the year 2022 for his significant contributions in the field of IT/ CSE. Dr. Barnwal has led several R&D and industrial consultancy projects funded by different agencies of Govt. of India in the field of Artificial Intelligence, Internet of Things, Data Informatics, and Cyber-Physical Systems. He has authored several research papers, book chapters, and technical reports in SCI journals and International Conferences. He also has several patents, copyrights, and design registrations to his credit for his authored software and technological products. During the COVID-19 era, he developed and filed a patent on IoT-based IntelliMAST (Solar-powered Intelligent Mask ATM cum Thermal Scanner) technology for the workplace, which attracted the attention of national/ international media. He is also associated with the Academy of Scientific and Innovative Research (AcSIR), India as its Associate Professor. Further to this, Dr. Barnwal has also served as one of the guest editors of the International Journal of Distributed Sensor Networks and is currently in the reviewers' panel of several high-impact-factor SCI journals. Recently, he has been awarded the prestigious INSA Visiting Scientist Fellowship for the year 2022-23 from Indian National Science Academy. His current areas of research interest include Intelligent Machines, Internet of Things, Cloud Computing, and Cyber-Physical Systems.