Pranit Gopaldas Shah
MEMORY STRUCTURES, SOFTWARE ENGINEERING, PROGRAMMING LANGUAGES, MODELS AND PRINCIPLES, INFORMATION SYSTEMS APPLICATIONS, ARTIFICIAL INTELLIGENCE, COMPUTER GRAPHICS, IMAGE PROCESSING AND COMPUTER VISION, PATTERN RECOGNITION, SIMULATION AND MODELING, DOCUMENT AND TEXT PROCESSING, ADMINISTRATIVE DATA PROCESSING, LIFE AND MEDICAL SCIENCES computer vision end-to-end learning artificial intelligence image processing pattern recognition simulation and modeling Computer Vision and Pattern Recognition

Bio

Pranit Gopaldas Shah is a dedicated researcher and academic professional affiliated with the Parul Institute of Engineering and Technology in Vadodara, India. With a strong foundation in Computer Science and Engineering, he holds an MTech in Computer Engineering, a BECE, an MPM, and the prestigious FCSRC designation. As the Chief Research Scientist at TeerHub Technology Private Limited, he has contributed to diverse areas including artificial intelligence, machine learning, computer vision, and deep learning. His authored work, 'Activation Function: Key to Cloning from Human Learning to Deep Learning,' reflects his deep interest in bridging human cognition with computational models. Beyond his own research, Pranit has served extensively as a reviewer and editor for the Global Journal of Computer Science and Technology, overseeing the publication of dozens of papers across a wide range of computer science disciplines. His expertise spans memory structures, software engineering, programming languages, information systems, image processing, and pattern recognition, making him a versatile contributor to the academic community.

Educational Journey

Parul University

Master of Technology in Computer Engineering, Master in Advanced Project Management, Mechanical Engineering • Computer Engineering

Parul University

MTech CE, BECE, MPM, FCSRC, Master of Technology in Computer Engineering

Experience

Parul University

0 - 0 • Computer Science and Engineering

Research

Activation Function: Key to Cloning from Human Learning to Deep Learning

Article June 8, 2020

Maneuvering a steady on-road obstacle at high speed involves taking multiple decisions in split seconds. An inaccurate decision may result in crash. One of the key decision that needs to be taken is can the on-road steady obstacle be surpassed. The model learns to clone the drivers behavior of maneuvering a non-surpass-able obstacle and pass through a surpass-able obstacle. No data with labels of “surpass-able” and “non-surpass-able” was provided during training. We have development an array of test cases to verify the robustness of CNN models used in autonomous driving. Experimenting between activation functions and dropouts the model achieves an accuracy of 87.33% and run time of 4478 seconds with input of only 4881 images (training + testing). The model is trained for limited on-road steady obstacles. This paper provides a unique method to verify the robustness of CNN models for obstacle mitigation in autonomous vehicles.