DETECTION AND CONVERSION OF VOICE TO TEXT USING A DISPLAY FOR HEARING IMPAIRED PERSON

Abstract

The hearing impaired can visually analyze the speech of the normal person. Hence they use the sign language or lip reading to understand the speech. There are various devices designed especially for them like that of the hearing aid, visual display. To provide a speech to text recognition for the hearing impaired person to make them visible what they cannot hear. Which is done through a display unit with the help of the raspberry pi processor along with a rasbian operating system to make the speech as text using speech to text recognition software.Which makes the hearing impaired person to visualize the voice as text .which done by de-noising. De-noising is any signal processing method which reconstruct a signal from a noisy one. Its goal is to remove noise and preserve useful information. So that the keywords which are to be viewed alone is been display over a display unit on a processor. The speech to text recognition is embedded over a hardware unit.

Specification

Description:Proposed System
In this section we demonstrate the proposed system, which include four different modules and their
functions. Firstly the audio been received from the normal person is being preprocessed using the de-noising
technique.The preprocessed audio is converted into speech to text which will help the hearing impaired to
visually view the preprocessed audio as text. The text is displayed over a display screen which is attached over
a processor. Each block consists of the steps related to each module.
MODULES
 Sound acquisition.
 De-noising.
 Speech to text recognition.
 Embedded over a hardware.
1.1 Sound acquisition
Noise is sound that is not required by the perceiver, because it is disagreeable, loud, or interferes with
hearing. This results in the subjective discretion between sound and noise, where any sound may be
considered noise depending on the perceiver. Speech acquisition or initial language acquisition focuses on the
development of spoken language person. While grammatical and syntactic knowledge can be seen as a part
of language acquisition, speech acquisition emphases on the development of speech perception and speech
production over the first years of a person lifetime.
1.2 Pre-processing(De-noising)
De-noising is any signal processing method which reconstruct a signal from a noisy one. Its goal is to
remove noise and reservation useful information. Noise reduction is the process of eliminating noise from
a signal. All recording devices, both analog and digital, have traits that make them susceptible to noise. Noise
can be random or white noise with no coherence or coherent noise introduced by the device mechanism or
processing algorithms. In electronic recording device, a major form of noise caused by random electrons that
heavily influenced by heat, stray from their designated path.
These stray electrons impact the voltage of the output signal and thus create detectable noise. The
stage of de-noising or noise reduction, also stated to as improving of speech degraded by noise, aims to
improve the speech signals quality. The objective is to improve the intelligibility, a measure of how
comprehensible speech is. Noise corrupting speech signals can be grouped coarsely into the following 3
classes:
1. Microphone related noise
2. Electrical noise (e.g. electromagnetically induced or radiated noise)
3. Environmental noise
The first two types of noise can be simply compensated by training the speech recognizers on
corresponding noisy speech samples, but compensating the environmental noise is not that basic, due to its
high variability. The basic problem of noise reduction is to reduce the external noise without disturbing the
unvoiced and low-intensity noise-like components of the speech signal itself.
1.3 speech to text recognition
Speech recognition is the inter corrective sub field of computational linguistics which incorporates
knowledge and research in the linguistics, computer science, and electrical engineering fields to progress
methodologies and technologies that enables the recognition and translation of spoken language into text by
computers and computerized devices such as those categorized as clever technologies and robotics. It is also
known as "automatic speech recognition" (ASR), "computer speech recognition", or just "speech to text"
(STT).Some SR systems use "training" (also called "acceptance")where an individual orator reads text or
isolated vocabulary into the system. The system analyzes the person's specific voice and uses it to fine-tune
the recognition of that person's speech, resulting in increased exactness. Systems that do not use training are
called "speaker independent" systems. Systems that use training are called "speaker dependent". Speech
recognition applications contain voice user interfaces such as voice dialing call routing demotic appliance
control, search (e.g. find a podcast where specific words were spoken), simple data entry (e.g., entering a
credit card number), preparation of structured documents, speech-to-text processing (e.g., word
processors or emails), and aircraft.
1.4 Embedded over a hardware unit
Finally the preprocessed speech which is converted into text as which is needed to the hearing
impaired. Which is been displayed over a display screen along with its processor. The hardware being used
here is the raspberry pi processor with a display unit. To make the hearing impaired to visualize the voice as
text. The hardware unit also has a vibrator which helps the hearing person to know when someone speech to
him/her the vibrator vibes. Which act as an alert to the hearing impaired person.
SUMMARY OF THE INVENTION
A hardware tool which is used to display the voice as text over the display unit attached on the hardware.
which helps the hearing impaired person to visually view the voice content as text once the voice has been
received by the speech to text recognition. The device been used is feasible to carryout by the hearing
impaired person.so that the hearing impaired person can also respond to the audio which they cannot hear is
been viewed as text over the hardware.
Fig 1.1 Architecture
In this section we demonstrate the proposed system, which include four different modules and their
functions. Firstly the audio been received from the normal person is being preprocessed using the denoising
technique.The preprocessed audio is converted into speech to text which will help the
hearing impaired to visually view the preprocessed audio as text. The text is displayed over a display
screen which is attached over a processor.fig[1.1] shows the overall architecture of the proposed
system. Each block consists of the steps related to each module.
Performance Evaluation
Fig1.2 performance evaluation
In fig 1.2 the performance is been evaluated based on the criteria. Which indicates the proposed work
is satisfactory compared to the existing speech to text recognition for people with normal hearing as
it achieved higher mean satisfaction score.
Fig 1.3 flow diagram
In above figure 4.1 the audio been received from the normal person is being preprocessed using the de-noising
techniques with the help of matlab.The preprocessed audio is converted into speech to text which will help the
hearing impaired to visually view the preprocessed audio as text. The text is displayed over a display screen
which is attached over a processor. , Claims:This project is for a social cause for the hearing impaired persons to make them view the
speech, what they cannot hear.
2. A hearing impairment is a hearing loss that prevents a person from totally receiving sounds
through the ear.
3. Since the hearing impaired person cannot receive the audio this project helps them to
visualize the audio what they cannot hear.
4. Detecting and converting the speech into text and to use a display unit along with its
processor.
5. The audio been received from the normal person is being preprocessed using the de-noising
techniques with the help of matlab .
6. The preprocessed audio is converted into speech to text which will help the hearing impaired
to visually view the preprocessed audio as text.
7. The text is displayed over a display screen which is attached over a processor.

Documents