Scribd
Upload
72 views5 pages

Raspberry Pi-Based Ai System For Speech Transcription

The project proposal outlines the development of a Raspberry Pi-based AI system for speech transcription, aiming to create a cost-effective and portable solution for converting spoken language into text. It highlights the advantages of using Raspberry Pi over traditional cloud-based systems, including lower costs, offline functionality, and enhanced adaptability to various languages and accents. The project includes specifications for both hardware and software requirements necessary for implementation.

Uploaded by

PRABHU G
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
72 views5 pages

Raspberry Pi-Based Ai System For Speech Transcription

The project proposal outlines the development of a Raspberry Pi-based AI system for speech transcription, aiming to create a cost-effective and portable solution for converting spoken language into text. It highlights the advantages of using Raspberry Pi over traditional cloud-based systems, including lower costs, offline functionality, and enhanced adaptability to various languages and accents. The project includes specifications for both hardware and software requirements necessary for implementation.

Uploaded by

PRABHU G
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

SRI DURGADEVI POLYTECHNIC COLLEGE

DEPARTMENT OF COMPUTER ENGINEERING

PROJECT PROPOSAL FOR THE YEAR 2024-25

TITLE OF PROJECT:

RASPBERRY PI-BASED AI SYSTEM FOR


SPEECH TRANSCRIPTION

PROJECT MEMBERS:

REGISTER NUMBER NAME OF THE STUDENT


23505337 BALARAMAN K

23505340 DHIVAKAR G

23505367 NITHISH KUMAR R

23505369 PAGADALA MOHITH SAI VENKAT

23591348 RAMAKRISHNAN M

23591349 VIDHESHAN R

GUIDED BY
Mr G.PRABHU B.E.,
LECTURE/CE

APPROXIMATE COST OF THE PROJECT: Rs. 10000 /-


RASPBERRY PI-BASED AI SYSTEM FOR
SPEECH TRANSCRIPTION

ABSTRACT

The project, Speech Transcription Using AI with Raspberry Pi, aims to develop a
cost-effective and portable system for transcribing spoken language into text. Utilizing the
Raspberry Pi as a processing unit, the system leverages advanced AI techniques to
accurately capture and transcribe speech in real-time. This project intends to address the
growing need for accessible speech recognition technology in various applications, such as
assistive devices, hands-free control systems, and data entry tools.

Previously, speech transcription systems primarily relied on cloud-based services or


dedicated hardware with substantial processing power, which limited their accessibility and
flexibility. Traditional systems often used technologies like Hidden Markov Models
(HMMs) and Gaussian Mixture Models (GMMs) to interpret and transcribe speech. These
techniques, while effective, required extensive computing resources and lacked adaptability
to diverse languages and accents. As a result, earlier speech recognition solutions were
largely confined to environments with strong internet connectivity and high-performance
computing infrastructures.

Audio transcription is useful for recording meeting minutes, converting lectures to


text, and various other applications. This device can also aid journalists in transcribing
interviews. Here, GPT and a microphone are employed for transcription.

In contrast, Raspberry Pi Zero is utilised for speech recognition, a form of artificial


intelligence programming. Understanding and reproducing phonetics requires significant
training and refinement. However, some free open-source software, including Google-STT,
can be used for this purpose.
RASPBERRY PI-BASED AI SYSTEM FOR

SPEECH TRANSCRIPTION

BLOCK DIAGRAM:

POWER
SUPPLY

USB RASPBERRY LAPTOP


MICROPHONEEO (DISPLAY)
PHONE PI 4

 Broadcom BCM2711,
 LPDDR4-3200
SDRAM
SD CARD  2.4 GHz and 5.0 GHz
IEEE 802.11ac AMPLIFIER
(RASBIAN OS)
wireless, Bluetooth
5.0, BLE
 Gigabit Ethernet
 2 USB 3.0 ports

SPEAKER
EXISTING AND PROPOSED SYSTEM:

EXISTING SYSTEM PROPOSED SYSTEM

 Typically requires powerful servers or  Utilizes a Raspberry Pi, a low-cost, portable


dedicated hardware with high processing microcontroller.
power.  Employs AI and machine learning techniques.
 Often based on cloud computing services  Designed to operate offline, making it
or dedicated devices. suitable for remote areas without internet
 requiring a stable internet connection access.
 High costs due to the need for expensive  Low-cost due to the affordable Raspberry Pi
hardware and cloud services. and open-source software.
 Limited adaptability to different  High accuracy with the capability to adapt to
languages, accents, and environments. diverse accents and languages through
 Data is often transmitted to cloud servers, advanced deep learning models.
raising potential privacy concerns.  Local processing on the Raspberry Pi ensures
 Higher power consumption due to the data privacy as no external servers are
need for powerful processing units. involved.
 Lower power consumption, making it more
energy-efficient and suitable for portable,
battery-operated use.
SPECIFICATION:

SOFTWARE REQUIREMENTS:

1. Python pip module - Pocketsphinx

2. Language ML Model - Spchcat

3. Rasbian Pi OS

HARDWARE REQUIREMENTS:

1. Raspberry Pi Board - IV, Broadcom BCM2711, SDRAM

2. Amplifier - PAM8403

3. Speech Processing Sensor - 2V, Microphone (-32dB ± 1dB) sensitivity

4. Speaker - 5V, 8 ohm, 0.5 Watts

5. Memory Card - CLASS 10, 1GB.

6. Power Supply - 5V, DC

7. Laptop (Display) - Intel i3, 4 GB RAM

You might also like