Maya to Spanish voice recognition IA

chevron-icon
Back
project-presentation-img
pedrogezn
Project Owner

Maya to Spanish voice recognition IA

Funding Requested

$61,000 USD

Expert Review
Star Filled Image Star Filled Image Star Filled Image Star Filled Image Star Filled Image 0
Community
Star Filled Image Star Filled Image Star Filled Image Star Filled Image Star Filled Image 3.1 (7)

Overview

We want to create an AI model to translate spoken Maya into written Maya and Spanish to empower Maya speakers and tourist on the mayan region. This project also Meets Mexico's Indigenous Language Law by helping to execute the language rights for minorities and targets the vast education market. Thus, big companies and public institutions might use the Model to offer services for Mayas in their native language as per Mexican law regulations. This will also increate the use of Singularity IA and open the door for creating models for other minoritarie languages in Mexico and other regions.

Proposal Description

How Our Project Will Contribute To The Growth Of The Decentralized AI Platform

We are aiming to target the Mexican Goberment and educational system, to use it to revitalize Maya  that will become a mayor client on the platform, that will become also in a promotor throguth the  INALI (National Institute of Indigenous Languages) that has the direction by law to promote all indigenous language initiatives.

Our Team

Israel Romero  Machine Learning Developer with a vast experience with Blockchain.

Pedro Teran   Principal Software Developer and languages enthusiast.  with more than 10 years of development experiences

Milenka Escobar.   project CEO,  A Master in Cultural managment with a lot of experiences leading tech projects.
Denise Antonio,    Especialist in Digital Marketing and Web 3.0 Marketing.

View Team

AI services (New or Existing)

 Mayan Voice to  Mayan text.

Type

New AI service

Purpose

Convert mayan voice to Mayan text

AI inputs

Mayan Audio

AI outputs

Mayan Text

 Mayan Voice to Spanish Text.

Type

New AI service

Purpose

Convert Mayan audio to Spanish Text

AI inputs

Mayan audio input

AI outputs

Spanish plain text

Mayan Text to Spansih text.

Type

New AI service

Purpose

Translate text from Maya to Spanish

AI inputs

Mayan Text

AI outputs

Spanish Text

Spanish Text to Mayan Text.

Type

New AI service

Purpose

Translate Spanish Text to Mayan Text.

AI inputs

Spansih Text

AI outputs

Mayan Text

The core problem we are aiming to solve

there are no Voice to text recognition Models for Maya language and also for other Indigenous langages that have a grammar conceptually different from Spanish and other languages because it uses demostratives that add sufixes to words  that make translations different,  which makes traditional IA models fail while tunning to Maya.

Maya is spoken by  around 795.000 mexicans. and encoureging them to keep using their language is a challenge in a world were they need to learn spanish to be able to work

Our specific solution to this problem

We want to create an IA Model,  to recognize Mayan voice, and translate it to Mayan text, and also translate it to Spanish Text, that we want to publish on Singularity IA.
To acomplish this, we will work with 2 linguist to create all the data sets.  and later with 3 mayan speaking  people to record  audios in different scenarios to build the data set to train the IA Model in collaboration with our DataScience team,  to train the model, and create the especific gramatical rules that  maya needs to be translated.


Project details

Objectives
Help Mayan people get included on goverment services

Market Size
Just in Mexico there are 1.569.047 public employees that potentially might use the services to translate documents, marketing and other resources to Maya as part of being compliant with the "indegenous lanugages rights act from 2005".
plus Mexico has 255,589 Schools that are potential users of the solution.

Value proposition. 

We are the only Mayan translation services that recognizes mayan voice
Comunication channels

b2b meeting and local marketing 

Delivery channels

SDNET platform

key Partners

Singularity IA, INALI (national institute for indegenous languages in mexico)

investments

59k usd to build and train the model

Core Business

sell the IA model as a service charging a fee to execute the model

Income sources

Business clients like mexican goverment, some big companies like banks and schools in Mexico

Costs and operations

Competition and USPs

Our Model to recognize Maya and translate it doesn't exist yet in the market that gives us the advantage, and also we will leverage the "indigenous languages rights act from 2005" to go sell our solution to the  goverment (we already talked about it), so they can implement our model in their process and solutions making the mexican Goverment our Main client, and from that on we want to start offering a solution to big companies like banks in Mexico to next start marketing Guatemala and el Salvador which also speak Maya. 
The main reason why we want to do it on Singularity IA is that we want to keep our model free from goverment influence, to be more aligned with comunity needs.

Open Source Licensing

GNU GPL - GNU General Public License

everything will be under GPL License

Revenue Sharing Model

API Calls

API Description:

we will publish 4 services  on the platform

  1.  Mayan Voice to  Mayan text.
  2.  Mayan Voice to Spanish Text.
  3.  Mayan Text to Spansih text.
  4. Spanish Text to Mayan Text.

API Revenue Service

1300

API Revenue Percentage

30

API Revenue Year

2026

Proposal Video

Placeholder for Spotlight Day Pitch-presentations. Video's will be added by the DF team when available.

  • Total Milestones

    5

  • Total Budget

    $61,000 USD

  • Last Updated

    18 May 2024

Milestone 1 - API Calls & Hostings

Description

This milestone represents the required reservation of 25% of your total requested budget for API calls or hosting costs. Because it is required we have prefilled it for you and it cannot be removed or adapted.

Deliverables

You can use this amount for payment of API calls on our platform. Use it to call other services or use it as a marketing instrument to have other parties try out your service. Alternatively you can use it to pay for hosting and computing costs.

Budget

$15,250 USD

Milestone 2 - Creation of the DataSet and Future Engineering

Description

data collection data cleaning data labeling

Deliverables

The dataset base to train the models.

Budget

$32,250 USD

Milestone 3 - Model and Training

Description

Selection and training of the models to recognize maya voice and convert it to text and also translate it to Spanish and vice Versa

Deliverables

Beta versions of the IA Models for the 4 services.

Budget

$9,000 USD

Milestone 4 - Model evaluation fine tuning and deployment.

Description

Evaluate and fine tune models and prepare them to deploy on Singularity IA SDNET

Deliverables

4 Services deployed to Singularity IA

Budget

$3,000 USD

Milestone 5 - Create a Landing Page for the project

Description

We need a landing page to promote and sell the API so we can showcase the use of the API to potential clients

Deliverables

A landing page where we can demo our Model to potential clients. like the mexican goverment

Budget

$1,500 USD

Join the Discussion (1)

Sort by

1 Comment
  • 0
    commentator-avatar
    Gombilla
    Jun 5, 2024 | 1:37 PM

    Hi there. I would like to comment that collecting a diverse range of audio recordings in different scenarios is crucial for training the AI model effectively. However, obtaining sufficient and representative data may be challenging, especially if there are limitations on access to native Mayan speakers or if certain dialects or variations of the language are underrepresented in the dataset. Your team may have to put this under consideration as well. Thanks !

Reviews & Rating

Sort by

7 ratings
  • 0
    user-icon
    CLEMENT
    Jun 5, 2024 | 1:39 PM

    Overall

    4

    • Feasibility 3
    • Viability 3
    • Desirabilty 4
    • Usefulness 4
    Preserves the Mayan linguistic identity.

    I would begin by saying that developing an AI model for recognizing and translating the Mayan language contributes to the preservation and promotion of indigenous languages and cultures. By providing tools for Mayan speakers to communicate and interact in their native language, the project supports linguistic diversity and cultural heritage preservation efforts.

    In the same vein, making Mayan language recognition and translation accessible through AI technology enhances accessibility and inclusivity for Mayan speakers, especially in digital communication and information exchange. This project will empower individuals within the Mayan community to participate more fully in modern society while preserving their linguistic identity.

    Kudos to the team !

  • 0
    user-icon
    Nicolad2008
    Jun 8, 2024 | 11:08 AM

    Overall

    3

    • Feasibility 2
    • Viability 2
    • Desirabilty 3
    • Usefulness 3
    Maya voice translation

    The Maya community support project has great potential to help them communicate with visitors and obey the indigenous Language Law of Mexico. The creation of AI models to translate Maya voices into Maya and Spain text can open the door for other minority languages ​​in Mexico and other areas, meeting legal needs, promoting teachers. Sex and language conservation. This project is likely to be implemented through cooperation with the Mexican government and the educational system to revive the Maya language, and access the vast educational market via Inali. It brings practical value by supporting the Maya people to use their language in communication and public services, encouraging the use and conservation of Maya language. However, the project must overcome the technical challenge because there is no voice recognition model for the Maya language and need an effective strategy to encourage the Maya to use their language in a multi -language world. .

  • 0
    user-icon
    TrucTrixie
    Jun 9, 2024 | 2:15 PM

    Overall

    3

    • Feasibility 3
    • Viability 3
    • Desirabilty 2
    • Usefulness 3
    What difficulties anticipated & What are solutions

    The Mayan language is part of the indigenous culture and language, everyone is very aware of this. If the team does this proposal well, it will also be considered an effort to preserve cultural heritage. Everything is still in the theoretical stage, when implemented in practice, more difficulties will arise. Hopefully the team can anticipate difficulties and come up with appropriate solutions.

  • 0
    user-icon
    Max1524
    Jun 9, 2024 | 2:36 AM

    Overall

    3

    • Feasibility 4
    • Viability 3
    • Desirabilty 3
    • Usefulness 3
    The appropriate proposal addresses language issue

    The problem remains that there is no speech-to-text recognition model for the Mayan language, other native languages ​​are facing a similar situation. So the speech recognition IA model will help overcome this problem. I expect this proposal to be successful soon because it is predicted to play a huge role in resolving the language barrier, even more so.

  • 0
    user-icon
    Tu Nguyen
    May 29, 2024 | 2:27 PM

    Overall

    3

    • Feasibility 4
    • Viability 3
    • Desirabilty 3
    • Usefulness 3
    Maya To Spanish Voice Recognition IA

    The problem that this proposal will solve is: there are no voice to text recognition models for Maya language and also for other Indigenous langages that have a grammar conceptually different from Spanish and other languages ​​because it uses demostratives that add sufixes to words that make translations different, which makes traditional IA models fail while tunning to Maya. The solution of this proposal is: they will create an IA model to recognize Mayan speech and translate it to Maya text, and also translate it to Spanish Text which we want to publish on Singularity IA.
    They should present more details about the milestones. They should divide the work into smaller parts. For example, what tasks are included in milestone 2 of data collection?

  • 0
    user-icon
    BlackCoffee
    Jun 10, 2024 | 1:13 AM

    Overall

    3

    • Feasibility 3
    • Viability 2
    • Desirabilty 3
    • Usefulness 3
    Milestones should present for easier understanding

    The milestones - specifically the 5 milestones presented are quite brief and I think are not enough to well demonstrate the feasibility of this proposal. I recommend that the presentation be more specific, easier to understand for readers, and include a duration for each milestone. Thank you team.

  • 0
    user-icon
    Joseph Gastoni
    May 19, 2024 | 8:28 AM

    Overall

    3

    • Feasibility 3
    • Viability 3
    • Desirabilty 3
    • Usefulness 4
    It has potential but requires careful planning

    The Mayan voice-to-text translation project has potential but requires careful planning and execution. Focusing on user needs, collaboration with Mayan communities, and clear metrics for success can increase the project's value. While the lack of existing competition is an advantage, it's important to consider how the project will compete with established translation services that might adapt to include Mayan.

    This project proposes an AI model to translate spoken Mayan into written Mayan and Spanish. Here's a breakdown of its strengths and weaknesses:

    Feasibility:

    • Moderate-High: Existing AI models can be adapted for speech recognition and translation, but require linguistic expertise for Mayan.
    • Strengths: The project leverages existing technologies and focuses on a specific, well-defined task.
    • Weaknesses: Gathering sufficient training data in spoken Mayan and addressing complex Mayan grammar might pose challenges.

    Viability:

    • Moderate: Success depends on securing funding, acquiring linguistic expertise, and demonstrating the model's accuracy and usefulness.
    • Strengths: The project aligns with Mexican law and potentially targets a large market (government, education).
    • Weaknesses: Competition from existing translation services (even if not voice-based) and limited initial user base for Mayan translation need to be considered.

    Desirability:

    • Moderate-High: Mayan communities and institutions seeking to comply with language rights could find this valuable.
    • Strengths: Empowering Mayan speakers and promoting cultural preservation are strong selling points.
    • Weaknesses: User needs and preferences for translation tools (e.g., mobile apps, web interfaces) need to be understood.

    Usefulness:

    • Moderate-High: The project can be very useful if it delivers accurate and user-friendly translation tools.
    • Strengths: Improved communication and access to services for Mayan speakers can have significant social benefits.
    • Weaknesses: Long-term impact on language revitalization and user adoption of the AI model requires evaluation.

    Besides, the project should consider:

    • Collaboration with Mayan linguists and communities is crucial for data collection, model development, and user testing.
    • Focusing on a mobile app or user-friendly interface can increase accessibility and adoption.
    • Demonstrating the model's accuracy and addressing any potential bias in translations is important for user trust.

    Here are some strengths of this project:

    • Addresses a critical need for Mayan language revitalization and promotes cultural preservation.
    • Leverages existing AI technologies for speech recognition and translation with a focus on a specific language.
    • Targets a potentially large market (government, education) aligned with Mexican law on indigenous languages.

    Here are some challenges to address:

    • Gathering sufficient and diverse training data for spoken Mayan and addressing the complexities of Mayan grammar.
    • Securing funding, acquiring linguistic expertise, and demonstrating the model's accuracy and usefulness compared to existing translation services.
    • Understanding user needs and preferences for translation tools (mobile apps, web interfaces) and ensuring user adoption within the Mayan community.

     

    user-icon
    pedrogezn
    May 20, 2024 | 8:28 PM
    Project Owner

    Hi Joseph, 

    thanks for the valuable feedback I have some things I would like to add regarding the points you mentioned.

    1 We know how to tackle the dataset Creation with mayan linguists already and is part of the budget hiring some. I been part of the  gogole-quechua model for google translate back then in Bolivia, so I know how to build a dataset and what we need from Mayan speaking people.

    2 Regarding an app or web to trnaslate, We are already in conversations with the Mexican government to buy this potential solution and integrate it in their services,  and from them we will focus on building this on an App, First we want to focus on making the model sustainable over time.

    3 regarding having a tool web or app to do the translations is something that we actually need is just that we didn't evaluated yet how to make it in a sustainable model while keeping it  free are at least cheap anough for Mayans to be able to afford it.  that's something that we will look further, for this project thanks for your thoughtfull review

Summary

Overall Community

3.1

from 7 reviews
  • 5
    0
  • 4
    1
  • 3
    6
  • 2
    0
  • 1
    0

Feasibility

3.1

from 7 reviews

Viability

2.7

from 7 reviews

Desirabilty

3

from 7 reviews

Usefulness

3.3

from 7 reviews

Get Involved

Contribute your talents by joining your dream team and project. Visit the job board at Freelance DAO for opportunites today!

View Job Board