Mahesh Jangid

About

Mahesh Jangid is a leading researcher in the field of document image analysis and pattern recognition, with a primary focus on handwritten character recognition for Indic scripts. His most impactful work, "Handwritten Devanagari Character Recognition Using Layer-Wise Training of Deep Convolutional Neural Networks and Adaptive Gradient Methods" (2018, 100 citations), addresses a critical challenge in assistive technology. By pioneering the use of layer-wise training of deep CNNs combined with adaptive gradient methods, Jangid significantly improved recognition accuracy for Devanagari script—a writing system used by over 500 million people. This contribution has direct implications for developing assistive tools for blind and visually impaired users, as well as advancing human–robot interaction and automated data entry systems. His research bridges the gap between deep learning methodologies and real-world applications in multilingual document processing. Jangid’s work has been widely cited by scholars working on OCR systems for low-resource languages, and his methodological innovations continue to influence the design of robust character recognition frameworks. Through his dedication to making written communication more accessible, Jangid has established himself as a key figure in the intersection of computer vision and assistive technology.

Research Focus

Key Achievements

Top Papers

1. 1
   Handwritten Devanagari Character Recognition Using Layer-Wise Training of Deep Convolutional Neural Networks and Adaptive Gradient Methods

   100 citations · 2018

Key Collaborators

Contact & Links