Prateek Keserwani

Prateek Keserwani

Ph.D. Scholar
Department of Computer Science and Engineering
IIT Roorkee

Research Area

  • Computer Vision
  • Machine Learning
  • Deep Learning

Contact Information

pkeserwani@cs.iitr.ac.in

Profile

Prateek Keserwani is a Research Scholar in Dept. of CSE, IITR, advised by Dr. Partha Pratim Roy (IIT Roorkee). He is interested in scene text understanding and exploring how text can be recognized and localized in scene images by using deep learning algorithms. He has completed M.Tech from the University of Allahabad, Allahabad (UP) in 2015. He has also served as a guest faculty in NIT Allahabad (from 2010 to 2012) and University of Allahabad (2015 - one semester).

The text appears on the captured natural images or videos is known as a scene text. The detection and recognition of the scene text is a challenging and important problem to address. It has ample applications such as robot navigation, an assistance system for visually impaired people, etc. The robust system to detect and recognition of scene text provides semantic information of the image/video which becomes helpful for better scene understanding. This is an important problem for the country such as India which have a multilingual society. The automatic reading system such an environment becomes very helpful to society. In our laboratory, we are developing the robust scene text detection system which covers the case of the multi-oriented, multi-lingual and curved text.

 

Fellowship

  • VVS-Fellowship for Ph.D. from the Ministry of Electronics and Information Technology (MeitY)
  • MHRD fellowship for M.Tech

 

Internship

  • 2 month internship in Indian Statistical Institute Kolkata under Prof. Umapada Pal.

 

Updates

  • Github link provided for the ICDAR 2019 challenge on Historical Document Reading Challenge on Large Structured Chinese Family Records for layout analysis task (Github link )
  • A paper has been accepted in ICDAR 2019 entitled "Zero Shot Learning Based Script Identification in the wild".
  • Acheived 3rd rank in ICDAR 2019 challenge on Historical Document Reading Challenge on Large Structured Chinese Family Records for layout analysis task.