TAMILNADU STATE COUNCIL FOR SCIENCE AND
TECHNOLOGY
STUDENT PROJECT PROPOSAL
PROPOSAL FOR STUDENT PROJECT SCHEME
ON
GPT – BASED SMART ASSISTANT FOR MULTILINGUAL DUBBING
PROCESS : FROM SCRIPT TRANSLATION TO FINAL AUDIO INTEGRATION
Submitted to TNSCST
SECTOR: Engineering and Technology
(2023-2024)
Submitted by
KOWSALYA A (613021205028)
DEEPANA S (613021205004)
MOWNIKA M (613021205034)
VIMALI D (613021205060)
Project Guide
Dr P PRABAHARAN
HEAD OF THE DEPARTMENT / IT
Application No. : CSE- 3122
TAMIL NADU STATE COUNCIL FOR SCIENCE AND TECHNOLOGY
APPLICATION FOR STUDENT PROJECT PROPOSAL (2024 – 2025)
Project Title : GPT based Smart Assistant (Multilingual Dubbing Process: From Script
Translation to Final Audio Integration)
Student Details :
S.No Name of the Student Mobile No. Email_Id
1. Kowsalya A 9843525031 shinydora753152@gmail.com
2. Deepana S 6369465159 deepanabtechit@gmail.com
3. Mownika M 9790618815 mownika18.m@gmail.com
4. Vimali D 8807827889 vimali.d001@gmail.com
Name of the Guide : Dr P Prabaharan
Designation : Head of the department
Department (Full Form) : Information Technology
Mobile Number : 9952349788
Email : kprabume@gmail.com
Vivekanandha College of Technology for Women,
Name of the Institution with Address : Sathinaickenpalayam, Elayampalayam Village,
Kumaramangalam – 637 205
Has a similar project been carried out
in your Institution / elsewhere? : No
Course Studying : UG Engineering
Project Details : Attached
Declaration
This is to certify that above mentioned students in table are bonafide final year students of P.G.
Science / U.G. Engineering / P.G. Professional Courses of our Institution and it is also certified that
two copies of utilization certificate and final report along with seminar paper will be sent to the
Council after completion of the project by the end of May 2025.
Signature of the Guide Signature of HoD Signature of the Principal/Registrar/ Dean
(with seal)
INTRODUCTION
The GPT-based smart assistant project aims to revolutionize the multilingual dubbing
process by automating script translation, voiceover generation, and audio synchronization
using cutting-edge Generative Pre-trained Transformer (GPT) technology. This innovative
solution will enable content creators to efficiently produce high-quality, multilingual audio
content, saving time and resources while expanding their global reach.
By leveraging GPT models, this smart assistant will:
Accurately translate scripts into multiple languages
Generate natural-sounding voiceovers
Synchronize audio with visual content
This project will transform the dubbing industry, enabling faster, more cost-effective,
and higher-quality content production for global audiences.
OBJECTIVES
Primary Objectives:
1. Automate script translation, voiceover generation, and audio synchronization for
multilingual dubbing.
2. Achieve high accuracy and quality in translated scripts, generated voiceovers, and
synchronized audio.
3. Reduce production time and costs associated with traditional dubbing processes.
Secondary Objectives:
1. Develop a user-friendly smart assistant interface for easy operation and monitoring.
2. Support multiple languages and dialects for global content creation.
3. Ensure scalability and flexibility for various content formats and industries.
4. Provide real-time feedback and quality control mechanisms.
Technical Objectives:
1. Train and fine-tune GPT models for script translation, voiceover generation, and audio
synchronization.
2. Develop and integrate necessary software components and APIs.
3. Ensure system reliability, security, and data privacy.
METHODOLOGY
Phase 1: Research and Planning
Phase 2: Data Collection and Preprocessing
Phase 3: GPT Model Training and Fine-Tuning
Phase 4: Engine Development (Translation ,Voice over , Audio Synchronization)
Phase 5: Integration and Testing
Phase 6: Deployment and Maintenance
WORKPLAN
Weeks 1-4: Research and Planning
Define project requirements and scope
Conduct literature review on GPT models and dubbing processes
Develop project timeline and resource allocation plan
Weeks 5-12: Script Translation
Collect and preprocess script datasets (Weeks 5-6)
Train and fine-tune GPT models for script translation (Weeks 7-8)
Develop translation engine and integrate with smart assistant (Weeks 9-12)
Weeks 13-20: Voiceover Generation
Collect and preprocess voiceover datasets (Weeks 13-14)
Train and fine-tune GPT models for voiceover generation (Weeks 15-16)
Develop voiceover engine and integrate with smart assistant (Weeks 17-20)
Weeks 21-24: Audio Synchronization
Collect and preprocess audio datasets (Weeks 21-22)
Train and fine-tune GPT models for audio synchronization (Weeks 23-24)
Weeks 25-30: Integration and Testing
Integrate script translation, voiceover generation, and audio synchronization engines
(Weeks 25-26)
Conduct comprehensive system testing and quality assurance (Weeks 27-30)
After Week 30: Deployment and Maintenance
Deploy smart assistant in cloud-based infrastructure
Provide ongoing maintenance, updates, and support
BUDGET
COMPONENT ESTIMATION COST(in Rs)
GPT Model Development 2500
Translation Module 800
Dubbing Software Integration 1000
Audio Engineering Tools 900
Project Management 550
Cloud Infrastructure 650
Total Cost 6400
THANKING YOU!
