Article Details
  • Published Online:
    April  2026
  • Product Name:
    The IUP Journal of Computer Sciences
  • Product Type:
    Article
  • Product Code:
    IJCS040126
  • DOI:
    10.71329/IUPJCS/2026.20.2.7-20
  • Author Name:
    Samarth Mishra, Devraj Mishra, Nikita Singh, Ashish Pandey and Shubham Kumar Singh
  • Availability:
    YES
  • Subject/Domain:
    Engineering
  • Download Format:
    PDF
  • Pages:
    7-20
Volume 20, Issue 2, April-June 2026
Breaking the Language Barrier: A Deep Learning-Powered Indian Sign Language Translator
Abstract

Language barriers significantly limit the scope of people suffering from hearing and speech impairment, hindering them from being able to communicate with the general community. Indian Sign Language (ISL), the only means of communication for the majority of this community in India, has no extensive digital translating capabilities. This paper proposes an ISL translator based on deep learning that can correctly interpret and transcribe hand signs to textual form, overcoming the communication barrier. The proposed system recognizes the ISL alphabet (A-Z) and numbers (1-9) correctly. Besides, the system includes edge detection and tracking landmarks, solving problems due to varying lighting, and partial occlusions. The system also develops an application over the web, so it has accessibility advantages. Through computer vision, coupled with the capabilities of deep learning, the study contributes to an inclusive digital communications platform, empowering people who depend on sign language.

Introduction

Communication, a basic task in human interaction, allows one to convey their thoughts, feelings, and ideas to others.