Multilingual Speech to Speech Translation Device
Our Final Year Project is about building a device using Raspberry Pi 3b+ for speech to speech translation. A lot of people who only know their native language face many difficulties when they go to foreign countries and want to connect with people there in different events like conferences, meetings
2025-06-28 16:34:13 - Adil Khan
Multilingual Speech to Speech Translation Device
Project Area of Specialization Computer ScienceProject SummaryOur Final Year Project is about building a device using Raspberry Pi 3b+ for speech to speech translation. A lot of people who only know their native language face many difficulties when they go to foreign countries and want to connect with people there in different events like conferences, meetings or for tourism purposes. So, the language barrier is always a huge issue for them. Foreigners who come to work in Pakistan face many difficulties while interacting with locals. For instance, Chinese people who are working here in Pakistan on different projects face difficulty in interaction as they only know their native language. So, they need a translation device that can bridge this communication gap.
Our device will provide more than 100 language pairs for translation. By using this device, first speech from user will be received by microphone. This speech will be recognized using ASR and converted into text. This text will then be translated into targeted language text using machine translation depending upon the user preference. Then targeted text will be synthesized using TTS into sound. For this we will use customized microphone and speaker compatible with raspberry pi.
There are multiple Speech to Speech translation devices like iFlytek, ENENCE Instant translator in the market that ease communication between people speaking different languages. Also, there are many apps like Google Translate, iTranslate, Microsoft Translator, etc. But the main drawback of the existing devices is that they do not provide support for the Urdu language, as all of them are made by foreigners. So, the motivation behind this project is to help Urdu speaking people in foreign countries to communicate with people there and erase the language barrier between locals and foreigners.
Project Objectives- To perform R&D for most suitable ASR, MT (Machine Translation) algorithms and TTS.
- To build a customized microphone and speaker compatible with raspberry pi 3b+ to provide speech interface.
- To design a portable, wireless device with its own powerhouse, screen and microphone using Raspberry pi3 as an embedded system.
- To solve an existing social problem using our engineering knowledge and cost-effective design solutions.
Complete architecture diagram of the system explaining how different modules relate to each other is shown below:
The ASR (Automatic Speech Recognition) module will be used to identify and process voice pf user. After processing, speech received from microphone will be converted into text. That text will then be translated from source language into target language by MT (Machine Translation) module. The text after being translated into target language is then passed to TTS (Text to Speech) module which will convert it into speech again.
The complete project consists of 4 phases. The first phase is configuration phase. In this phase all the hardware components will be configured with raspberry pi. These components include USB microphone, speaker and 3.5 inches LCD screen. Input Output audio ports of raspberry pi will also be configured to send/receive audio from microphone and speaker. Once the raspberry pi is ready to use with all the required hardware components, we can move onto second phase.
In the second phase, all the software work will be done. A virtual environment will be created on raspberry pi having all the dependencies and libraries installed for running the python program. A complete python program which receives input audio from mic, translates it into the desired target language and then converts that translated text to speech again will be written. Moreover, a complete code for displaying UI on the LCD screen will also be written in this phase. In the third phase, hardware and software will be integrated with each other. The python codes written in the second phase are made to run on raspberry pi.
In the fourth phase, final testing of the device is done to make sure there is no error left and the device is ready to use. We’ll be using agile methodology in all of the phases.
Benefits of the ProjectSome of the benefits of project are described below:
- The device will bridge the communication gap between people belonging to different countries and speaking different languages.
- The device will help tourists to communicate with locals and in business meetings and conferences to communicate with foreigners.
- As Urdu language is included in the list of available languages for translation, so this device will help local Pakistani workers to communicate with Chinese people working in Pakistan.
The main components associated with our project are:
- Raspberry pi3
- LCD 3.5 inch
- USB Microphone
- Power House
- Mini Speaker
A USB microphone will be the medium to provide voice signals to our program. The microphone is designed using the speaker used in mobile phones combined with an amplifier. We are using raspberry pi3 as a main tool in our project whose OS will operate the whole device using our python program. The python code will serve as translator software using online Google Cloud APIs for Machine Translation. This API allows us to use a million characters for 20$. The text will be converted into required language whether it is Urdu, English or any foreign language pair. The translated text will be shown on the LCD display. In the last step, the translated text will be converted back into speech. The raspberry needs at least a 2A current for charging and operating well. All the components are enclosed in a casing to make a portable device.
Final Deliverable of the Project HW/SW integrated systemCore Industry ITOther Industries Others Core Technology Artificial Intelligence(AI)Other Technologies 3D/4D PrintingSustainable Development Goals Decent Work and Economic Growth, Industry, Innovation and InfrastructureRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 22680 | |||
| Raspberry pi 3b+ | Equipment | 1 | 7000 | 7000 |
| Usb Mic | Equipment | 1 | 600 | 600 |
| Usb Speaker | Equipment | 1 | 800 | 800 |
| 3.5-inch LCD | Equipment | 1 | 3500 | 3500 |
| Memory Card | Equipment | 1 | 500 | 500 |
| Power Supply | Equipment | 1 | 4200 | 4200 |
| Device Casing | Equipment | 1 | 3000 | 3000 |
| Translation API price | Miscellaneous | 1 | 3080 | 3080 |