Natural Language Translation Engine for Announcements and Information Dissemination at Stations
The Smart India Hackathon (SIH) 2023 featured Problem Statement SIH1348, titled "Natural Language Translation Engine for Announcements and Information Dissemination at Stations." This challenge, presented by the Ministry of Railways, sought innovative solutions to enhance communication across India's diverse linguistic landscape.
Objective: Develop a robust Natural Language Processing (NLP) engine capable of translating station announcements and disseminating information into multiple regional languages. The solution should ensure accurate and timely translations to cater to the linguistic diversity of passengers.
Key Requirements:
- Translation Accuracy: The engine must provide precise translations, maintaining the context and intent of the original announcements.
- Real-Time Processing: Enable swift translation and dissemination to ensure passengers receive information promptly.
- Integration with Existing Systems: Seamlessly integrate with current station announcement systems and Information Dissemination Systems (IDS).
- Support for Multiple Languages: Cover a wide range of regional languages prevalent across different states and regions.
- User-Friendly Interface: Design an intuitive interface for operators to manage and monitor translations effectively.
- IVRS Technologies: Incorporate Interactive Voice Response System (IVRS) technologies to facilitate automated voice reminders in multiple languages for appropriate registered passengers.
| Category | Technologies |
|---|---|
| Frontend |
ReactJS
ThreeJS
TensorFlowJS
|
| Mobile Application |
Flutter
Dart
|
| Machine Learning | OpenAI Whisper-v3 | mbart-50 | Open Source LLMs |
| Backend |
NodeJS
PostgreSQL
|
|
Web |
App |
- Vedica Bafna (Team Leader)
- Paarshva Chitaliya
- Arjun Shah
- Hetansh Shah
- Siddhant Dutta
- Varun Viswanath