HACKVISION VAULT 2K25
TITLE PAGE
• Problem Statement ID – PSHVLT-036
• Problem Statement Title- AI-powered ISL Translator
• Theme- AI & Accessibility Innovation
• PS Category- Software
• Team ID- HVLT-036
• Team Name- TEAM ELITE
TEAM
ELITE
INDIAN SIGN LANGUAGE TRANSLATOR
Proposed Solution
• AI-powered real system for real-time ISL recognition and
translation.
• Converts ISL gestures into text or speech.
• Enables seamless integration with video and communication
platforms.
• Supports multiple devices: web, mobile and desktop.
2
TEAM
ELITE
TECHNICAL APPROACH
Frontend (React, WebRTC)
Computer Vision (OpenCV, Machine Learning CNNs, NLP & Speech
MediaPipe, YOLOv8, TF) (LSTM/RNN) (Gesture Recog.) (Transforemers, Google TTS)
Backend & API (Flask/FastAPI,
Firebase/MongoDB)
Cloud Integration (AWS/GCP)
3
TEAM
ELITE
FEASIBILITY AND VIABILITY
• Challenges: Real-time accuracy, dataset availability, scalability.
• Solution: Optimized AI models and efficient data processing.
• Scalability: Cloud based services ensures accessibility worldwide.
• User Adoption: Intuitive UI to encourage easy usage.
4
TEAM
ELITE
IMPACT AND BENEFITS
• Education: Enhances accessibility for online learning.
• Employment: Facilitates workplace communication.
• Public Services: Improve accessibility for government instructions.
• Scalability: languages Potential to expand beyond ISL to other sign
• Cost-Effective: Reduces the need for human translators.
5
TEAM
ELITE RESEARCH AND REFERENCES
• Academic papers on AI-driven sign language
recognition.
• Existing sign language datasets for model training.
• Accessibility guidelines and inclusive technology
research.
• WHO data on global hearing impairment and the
need for sign language accessibility.
• It is estimated that by 2050 over 700 million people
– or 1 in every 10 people – will have disabling
hearing loss