ZabanZad: Open-Source Persian Text-to-Speech AI
Donation protected
ZabanZad: Open-Source Persian Text-to-Speech AI
Introduction to the Project: As we venture into the future, technology is playing an increasingly integral part in bridging languages and cultures across the globe. Persian, a language that is often overlooked in the tech world, is speaking up, and we are standing at the forefront of this game-changing initiative: the ZabanZad Project. This project is led by the SAIL Lab @ University of New Haven, and is dedicated to transforming the Persian text-to-speech (TTS) landscape by developing open-source datasets and TTS models specifically for the Persian language and its diverse dialects.
Call to Action for Support: However, to achieve this grand vision, we need your help. Creating high-quality text-to-speech datasets is fundamental to the development of reliable speech synthesis models, and it's a resource-intensive task.
Our effort includes:
- Recruiting professional voice actors for studio-grade recordings.
- Curating diverse text sources for narration.
- Developing specialized recording software.
- Executing post-processing tasks for clarity and quality of datasets.
- Meticulously fine-tuning, training, and benchmarking our models.
Budget and Impact: The estimated total budget for this comprehensive project is $104,000. This amount will cover the meticulous and multifaceted tasks required to bring this initiative to life. With your donation, you're not just supporting a project; you're championing a cause dedicated to technological inclusivity and linguistic diversity.
Vision and Cultural Significance: Through the ZabanZad Project, we aim to establish the Persian language on equal footing with other languages in the digital communication landscape and uphold the cultural and linguistic diversity that is the heart of our global community. Your contribution can make a significant difference in the lives of millions.
Donation Tiers:
- $20 - Project Contributor: With a $20 donation, supporters play a vital role in the overall success of the project.
- $50 - Voice Actor Supporter: This level supports the recruitment of professional voice actors. Donors will help cover the costs of hiring native speakers for recording the datasets.
- $100 - Recording Innovator: Contributions at this level go towards the development of bespoke recording software. This software is essential for ensuring accurate text-to-speech matching and audio segmentation.
- $250 - Quality Assurance Champion: This funding level aids in the post-processing phase, ensuring the highest audio quality. It includes advanced denoising and audio editing.
- $500 - AI Model Pioneer: At the top level, donations contribute to the model training and benchmarking phase, using state-of-the-art models.
- $1000 - Technology Enabler: By joining the Technology Enabler tier with a $1,000 donation, you play a crucial role in advancing the technological backbone of the ZabanZad Project. Your contribution supports the acquisition and maintenance of high-performance computing resources essential for the development of sophisticated AI algorithms. This level of support is instrumental in enabling us to process large datasets and refine our text-to-speech models, ensuring they are both efficient and effective.
- $5000 - Infrastructure Architect: As an Infrastructure Architect with a generous donation of $5,000, you are at the heart of building a robust foundation for the ZabanZad Project. Your support is directed towards establishing a scalable and secure cloud infrastructure. This infrastructure is key to hosting our services, managing vast amounts of data, and providing seamless access to developers and users globally. Your contribution at this level is vital in laying the groundwork for a globally accessible Persian TTS service.
- $10,000 - Visionary Innovator: The Visionary Innovator tier, with a contribution of $10,000, is for those who wish to have a transformative impact on the future of the Persian language in the digital age. Your support at this level accelerates our research and development efforts, enabling groundbreaking innovations in Persian text-to-speech technology. It also supports our outreach and collaboration initiatives with leading tech companies and academic institutions. As a Visionary Innovator, you are not just funding a project; you are shaping the future of how Persian culture and language interact with cutting-edge technology.
Closing and Further Support: Stand with us as we strive not only to shape the future of Persian Text-to-Speech technologies but also to create a more inclusive and accessible space for Persian speakers worldwide. Please support our GoFundMe campaign, donate what you can, and help us to share our cause. Together, we can make this vision a reality. Furthermore, we aim to create real-world applications, such as a Persian-centric smart assistant and a web browser plugin for visually impaired Persian speakers. ZabanZad is a project by and for the Iranian diaspora. No datasets come from Iran and we will not be sourcing any funds from Iran.
Disclaimer: ZabanZad is a project by and for the Iranian diaspora. No datasets come from Iran and we will not be sourcing any funds from Iran.
Organizer
SAIL Lab University of New Haven
Organizer
West Haven, CT