Top Banner

of 24

Speech Banking System

Jun 03, 2018

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
  • 8/12/2019 Speech Banking System

    1/24

    INTRODUCTION

    SPPECH OPERATED bank lockersystem

    In todays world of technology every one think about comfort like controlling bank locker voice

    etc this system is Automatic recognition of speech by a machine has been a goal of research

    for more than four decades and has inspired technological wonders such as the ASIMO (a

    humanoid) AI!O (an android) robots by "O#$A%

    &ith life getting increasingly more demanding' many of us find our plans can change at a

    moments notice and have a need to be able to change settings in our home whether it is the

    heating system' telephone system' security system or the ability to choose not to start the

    washing machine or cooker at a previously chosen start time% Speech ecognition Machine give

    you these options and many more including the ability to see and speak to someone at your

    home front door whilst you are elsewhere%Our aim is to integrate speech recognition into the

    various consumer electronic devices%!ut an important uestion is' *&hat do we mean by

    speech recognition by Machine+,-oice or speech recognition is the ability of a machine or

    program to receive and interpret dictation' or to understand and carry out spoken commands%

    .articularly speech recognition machine is a system where the user speaks a command and the

    machine' upon correctly recogni/ing the command acts appropriately% 0hus the user e1ercises

    control over the machine using simple voice commands%

    PROPOSED SYSTEM

    O2 .O3450 $4A6S &I0" bank locker 5O#0O66I#7 S8S04M control system For severalyears, demonstration 'bank locker of the future' have shown control of bank locker

    devices by voice. There are some obvious benefits. Voice allows hands-free control.This is vital for people who find keyboards difficult or impossible to use, such system

    provide high security since we need to tarin the system with the users voice only'Spoken dialogue technology' takes voice control one step further, allowing a system to

    interact in a step-by-step fashion with a user, asking uestions and responding to the

  • 8/12/2019 Speech Banking System

    2/24

    replies. This makes advanced functionality of multiple, comple!, networked devicesmuch more accessible, not "ust to the technically-minded with the patience to wadethrough manuals. #t also allows people to phone up their home and access devicesremotely, for e!ample to turn on the heating or air conditioning before returning.

    Voice control and spoken dialogue is starting to become commonplace in cars, not "ust

    in top-of-range models. So what is the potential in the home, what are the componenttechnologies, and what is needed for it to become eually commonplace$

    The technology

    'Speech recognition' is the process of taking a speech signal and converting it intowords. The technology has improved steadily, if not spectacularly, over the last fewdecades, but it is still not possible to accurately convert anyone's voice talking aboutany sub"ect. The recogniser therefore needs to be trained to one or more specificspeakers, i.e. 'speaker-dependent, large vocabulary' recognition, or restricted in thenumber of words it can recognise, i.e. 'speaker independent, small vocabulary'.

    A general model for speech recognition is as shown%

    9igure :;:7eneral !lock diagram of speech recognition system%0he model begins with a user creating a speech signal (speaking) to accomplish a given task%

    0he spoken output is first recogni/ed' in that the speech signal is decoded into a series of words

    that are meaningful according to synta1' semantics and pragmatics according to the conte1t of

    what it has previously recogni/ed% 0he feedback from high level processing reduces the

    comple1ity of the recognition model by limiting the search for valid input speech from the user%

    Word

    Recognition

    Model

    High

    Leel

    Processing

    !oice

    O"t#"t

    Task

    Descri#tion

    Synta$

    Semantics

    Pragmatics

    Dynamic

    %no&ledge

    Re#resentation

    'ser

    S#eech

    'ser

    S#eech

  • 8/12/2019 Speech Banking System

    3/24

    0he recognition system responds to the user in the form of voice output or euivalently in form

    of reuested action being performed' with user being prompted for more input%

    One of the most difficult aspects of building up speech recognition machine is its

    interdisciplinary nature% 5onsider the discipline that have been applied to speech recognition%

    1. Signal processing< the process of e1tracting relevant information from the speech in an

    efficient' robust manner ;speech analysis and enhancement%

    2. .hysics (acoustics)< the science of understanding the relationship between the physical

    speech signal and the physiological mechanisms (the human vocal tract mechanism)

    that produced the speech and with which the speech is perceived %

    3. Information 0heory:statistical modeling of speech signals%

    =% 6inguistics< the relationships between sounds (phonology)' words in a language

    (synta1)' meaning of spoken words (semantics) and sense derived from meaning %

    >% .hysiology:understanding of the higher;order mechanisms within the human central

    nervous system that account for speech production and perception in human beings

    ;Artificial neural network%

    ?% 5omputer science:the study of efficient algorithms for implementing' in software or

    hardware' the various methods used in practical speech recognition systems

    @% .sychology:the science of understanding the factors that enables a technology to be

    used by human beings in practical tasks%

    "owever' in spite of the glamour of designing an intelligent machine that can recogni/e the

    spoken word and comprehend its meaning' we are far away from achieving the desired goal of

    a machine that can understand spoken discourse on any subect by all speakers in all

    environments%

    Chapter 2: LITERATURE SURVEY

    2.1 Speech Productio

    9igure B%: portrays a medium saggital section of the speech system in which we view the

    anatomy midway through the upper torso as we look on from the right side C:D% 0he gross

    components of the system are the lungs' trachea (windpipe)' laryn1 (organ of speech

    production)' pharyngeal cavity (throat)' oral or buccal cavity (mouth)' and nasal cavity (nose)% In

    technical discussions' the pharyngeal and oral cavities are usually grouped into one unit

  • 8/12/2019 Speech Banking System

    4/24

    referred to as the vocal tract' and the nasal cavity is often called the nasal tract% Accordingly' the

    vocal tract begins at the output of the laryn1 (vocal cords' or glottis) and terminates at the input

    to the lips% 0he nasal tract begins at the velum and ends at the nostrils%

    PROBLEM STATEMENT:

    banking a##lications like controlling the locker &ith high sec"rity this system &ill be ery m"ch

    #roen technology to control the bank locker by a single n"mber locker o#ening In todays world

    of technology every one think about comfort like controlling bank locker voice etc this system is

    Automatic recognition of speech by a machine has been a goal of research for more than four

    decades and has inspired technological wonders

    METHODOLOGY: 'sing s#eech control circ"it and microcontroller is the method to ado#t

    and to reach the goal o( accessing the locker

    !LOC" DIA#RA$

    5onsists of

    :% microphone

    B% voice processor

    E% microcontroller

    =% driver circuit

    >% lcd display system

    ?% power supply

    Micro phone is the one type of transducer which converts voice signal to electrical

    signal% 0hese electrical signals are very small mill voltage signal' so it is given to

    amplifier circuit% 0he amplifier circuit is constructed with operational amplifier which

    acts as power amplifier which is inbuilt in the voice recognition ic itself % 0hen the

    amplifier signal is given to signal conditioning unit which also constructed with

    operational amplifier% In this circuit operational amplifier act as comparator and

  • 8/12/2019 Speech Banking System

    5/24

    generate the suare pulse given to microcontroller% "ere the microcontroller may be

    Atmel or .I5 both are flash type reprogrammable microcontroller%

    In microcontroller we have already programmed so it receives the pulse signal

    from signal conditioning unit and activates the relay driver circuits% $river circuits

    consists of 0ransistor' it ust act as switch to turn O#' turn O99 the relays% elay

    outputs are directly connected to ant appliances or system or even motors which

    are attached in the robot%

    S&itches

    M)CRO

    CO*TROLLER

    Locker S+STEM

    POWER S'PPL+

    Oscillator

    )*D)CATORS,ROM !O)CE

    RECO-*)T)O*

    )C

  • 8/12/2019 Speech Banking System

    6/24

    #eera% de&criptio:

    "$BFF: is a monolithic user dependence

    speech recognition I5 designed for toy

    application% "$BFF: consist of microphone

    amplifier' AG$ converter' speech processor'

    and IGO controller% After pre;recording'

    "$BFF: can recogni/e up to @ different

    &ord sentences each with :%> sec long'%

    &ith highly IGO .rogrammability% "$BFF:

    can be adapted in a wide range application

    for toy products%

  • 8/12/2019 Speech Banking System

    7/24

  • 8/12/2019 Speech Banking System

    8/24

  • 8/12/2019 Speech Banking System

    9/24

    Itroductio to $icro'Cotro%%er ()C*1

    has gained great popularity since its introduction% Its standard form includes severalstandard on;chip peripherals' including timers' counters' and 2A0Hs' plus =kbytes ofon;chip program memory and :B@ bytes (note< bytes' not bytes) of data memory'making single;chip implementations possible% Its hundreds of derivatives' manufacturedby several different companies (like .hilips) include even more on;chip peripherals'such as analog;digital converters' pulse;width modulators' IB5 bus interfaces' etc%5osting only a few dollars per I5' the @F>: is estimated to be used in a largepercentage (maybe :GB+) all embedded system products%

    0he @F>: memory architecture includes :B@ bytes of data memory that are accessibledirectly by its instructions% A EB;byte segment of this :B@ byte memory block is bit

    addressable by a subset of the @F>: instructions' namely the bit;instructions% 41ternalmemory of up to ?= bytes is accessable by a special Jmov1J instruction% 2p to =bytes of program instructions can be stored in the internal memory of the @F>:' or the@F>: can be configured to use up to ?= bytes of e1ternal program memory 0hemaority of the @F>:Hs instructions are e1ecuted within :B clock cycles%

    &e developed a -"$6 synthesi/able model of the @F>: and a 5KK based @F>:instruction;set simulator' both found below' on which weHve based some researchdirections% One of those directions is a tuning environment' also found below' to assist adesigner who wants to modify the @F>: architecture to be more power efficient for aparticular program 8ou see' a particular @F>: will probably e1ecute the same program

    over and over for its lifetime' so it would be nice to orient the @F>: towards thatprogram% &eHve also developed some new architectural features that can be used tolower the power of an @F>:' which will appear on this page in the future%

    One of the defining characteristics of embedded systems is that they run oneapplication for their entire e1ecution time% 0his characteristic creates a great potentialfor optimi/ations% 9or e1ample' a general purpose processor might have a datapathconsisting of a register file' A62' and a shifter% 9or an instruction such as a move' this

    !O)CE

    RECO-*)T)O*

    )C

    Mike

    RAM

    . SE-EME*T

    D)SPLA+

    DR)!ER

    DR)!ER

    . SE-EME*T

    D)SPLA+

    POWER S'PPL+

  • 8/12/2019 Speech Banking System

    10/24

    datapath is e1pensive in terms of power because it must pass through the A62 andshifter before being written back into the register file% If this move operation was verycommon to a particular application' an optimi/ation could be made by modifying theregister file to move data from one location to another%

    0he information on this page describes an environment that simplifies the process oftuning a microprocessor% 0hree main tools are provided in order to analy/e the powerconsumption of an @F>: microprocessor% 0he first of these tools is an architectural viewwhich generates a hierarchical display of power consumption throughout thecomponents of the processor% 0he output is formatted so that it is readable by 1du' agraphical program for displaying tree structures% 0he second tool calculates the averagepower per instruction for the entire instruction set (several ump instructions have not yetbeen implemented in this tool)% 0he third tool is an instruction set simulator whichoutputs a program memory view that displays the instruction' power' and freuency forevery address% It also provides a data memory view that calculates accesses for eachlocation in ram%

    +eature&:

    5ompatible with M5S;>: .roducts%

    = !ytes of In;System eprogrammable 9lash Memory%

    ; 4ndurance< :FFF &riteG4rase 5ycles

    9ully Static Operation< F "/ to B= M"/

    0hree;level .rogram Memory 6ock%

    :B@ 1 @;bit Internal AM%

    EB .rogrammable IGO 6ines%

    0wo :?;bit 0imersG5ounters%

    Si1 Interrupt Sources% .rogrammable Serial 5hannel%

    6ow;power Idle and .ower;down Modes%

  • 8/12/2019 Speech Banking System

    11/24

    De&criptio:

    0he digital output of A$5 is given to the microcontroller (@L5>:)% 0he clock input is alsogiven to the microcontroller for the synchroni/ation% 0he A0@L5>: is low;power high

    performance @;bit microcomputer with = bytes of flash programmable and erasable read only

    memory (.4OM)% 0he device is manufactured using Atmels high density non volatile memory

    technology and is compatible with the industry standard M5S;>: instruction set and pin out% 0he

    on;flash allows the program memory to be reprogrammed in system or by a conventional non;

    volatile memory programmer% !y combining a versatile @;bit 5.2 with flash on a monolithic

    chip' the Atmel A0@L5>: is a powerful microcomputer which provides a highly fle1ible and cost

    effective solution to many embedded control application%

  • 8/12/2019 Speech Banking System

    12/24

    Pi De&criptio:

    VCC

    Supply voltage%

    #ND

    7round%

    Port ,:

    .ort F is an @;bit open;drain bi;directional IGO port% As an output port each pin can sink @

    006 inputs% &hen :s are written to port F pins' the pins can be used as high impedance inputs%

    .ort F may also be configured to be the multiple1ed low;order addressGdata bus duringaccesses to e1ternal program and data memory% In this mode .F has internal pull;ups% .ort F

    also receives the code bytes during 9lash .rogramming' and outputs the code bytes during

    program verification% 41ternal pull;ups are reuired during program verification%

    Port 1:

    .ort : is an @;bit bi;directional IGO port with internal pull;ups% 0he port : output buffers can

    sinkGsource = 006 inputs% &hen :s are written to .ort : pins they are pulled high by the internal

    pull;ups and can be used as inputs% As inputs' .ort : pins that are e1ternally being pulled by low

    will source current (II6) because of the internal pull;ups% .ort : also receives the low;order

    address bytes during 9lash .rogramming and verification%

  • 8/12/2019 Speech Banking System

    13/24

    Port 2:

    .ort B is an @;bit bi;directional IGO port with internal pull;ups% 0he .ort B output buffers can

    sinkGsource = 006 inputs% &hen :s are written to .ort B pins they are pulled high by the internal

    pull;ups and can be used as inputs% As inputs' .ort B pins that are e1ternally being pulled low

    will source current (II6) because of the internal pull;ups%

    .ort B emits the high;order address byte during fetches from e1ternal program memory

    and during accesses to e1ternal data memory that use :?;bit address (MO-N $.0)% In this

    application' it uses strong internal pull;ups when emitting :s% $uring accesses to e1ternal data

    memory that use @;bit addresses (MO-N :)% .ort B emits the contents of the .B Special

    9unction egister(S9)% .ort B also receives the high;order address bits and some control

    signals during 9lash .rogramming and verification%

    Port -:

    .ort E is an @;bit bi;directional IGO port with internal pull;ups% 0he .ort E output buffers can

    sinkGsource = 006 inputs% &hen :s are

    written to .ort E pins they are pulled high by the internal pull;ups and can be used as inputs% As

    inputs' port E pins that are e1ternally being pulled low will source current (I I6) because of the

    pull;ups% .ort E also serves the functions of the various features of the A0@L5>: as listed

    below 0: (timer : e1ternal input)

    .E%?

    PPP

    & (e1ternal data memory write strobe)

    .E%Q

    PPP

    $ (e1ternal data memory read strobe)

    RST:

    eset input% A high on this pin for two machine cycles while the oscillator is running resets

    the device%

    ALEPRO#:

    Address 6atch 4nable output pulse for latching the low byte of the address during

    accesses to e1ternal memory% 0his pin is also the program pulse input (.O7) during 9lash

    .rogramming% In normal operation A64 is emitted at a constant rate of :G? the oscillator

    freuency' and may be used for e1ternal timing or clocking purposes% #ote' however' that one

    A64 pulse is skipped during each access to e1ternal $ata Memory%

  • 8/12/2019 Speech Banking System

    15/24

    If desired' A64 operation can be disabled by setting bit F of S9 location @4"% &ith the bit

    set' A64 is active only during a MO-N or MO-5 instruction% Otherwise' the pin is weakly pulled

    high% Setting the A64;disable bit has no effect if the microcontroller is in e1ternal e1ecution

    mode%

    /////

    PSEN:

    .rogram Store 4nable is the read strobe to e1ternal program memory% &hen the

    A0@L5>: is e1ecuting code from e1ternal program memory' .S4# is activated twice eachmachine cycle e1cept that two .S4# activation are skipped during each access to e1ternal data

    memory%

    //

    EAVPP:

    PP

    41ternal Access 4nable% 4A must be strapped to 7#$ in order to enable the device tofetch code from e1ternal program memory locations starting at FFFF" up to 9999"% #ote

    however' that if lock bit;: is programmed' 4A will be internally latched on reset% 4A should be

    strapped to -55 for internal program e1ecutions% 0his pin also receives the :B- programming

    enable voltage (-..) during flash programming for parts that reuire :B-olt -..%

    0TAL1:

    Input to the inverting oscillator amplifier and input to the internal clock operating circuit%

    0TAL2:

    Output from the inverting oscillator amplifier%

  • 8/12/2019 Speech Banking System

    16/24

    Id%e $ode:

    In idle mode' the 5.2 puts itself to sleep while all the non;chip peripheral remain active%

    the mode is invoked by software% 0he content of the on;chip AM and all the S9 remain

    unchanged during this mode% 0he idle mode can be terminated by any enabled interrupt or by a

    hardware reset% It should be noted that when idle s terminated by a hardware reset the devicenormally resumes program

    e1ecutions' from where it left off' up to two machine cycles before the internal reset algorithm

    takes control% On;chip hardware inhibits access to internal AM in this event' but access to the

    port pins is not inhibited% 0o eliminate the possibility of an une1pected write to a port pin when

    idle is terminated by reset the instruction following the one that invokes idle should not be one

    that writes to a port pin or to an e1ternal memory%

  • 8/12/2019 Speech Banking System

    17/24

    Poer'do $ode:

    In the power;down mod' the oscillator is stopped' and the instruction that invokes power;

    down is the last instruction e1ecuted% 0he on;chip AM and S9 retain their values until the

    power;down mode is terminated% 0he only e1it from power;down is the hardware reset% esetredefines the S9s but does not change the on;chip AM% 0he reset should not be activated

    before -55is restored to its

    normal operating level and must be held active long enough to allow the oscillator to restart and

    stabili/e%

  • 8/12/2019 Speech Banking System

    18/24

    Prora3 $e3or4 Loc5 !it&:

    On the chip are three lock bits which can be left unprogrammed (2) or can be

    programmed (.) to obtain the additional features listed in the table below% &hen lock bit : is

    programmed' the logic level at the 4A pin is sampled and latched during reset% If the device is

    powered up without a reset' the latch initiali/es to a random value' and holds the value until

    reset is activated% It is necessary that the latched value of 4A be in agreement

    with the current logic level at that pin in order for the device for the device to function properly%

  • 8/12/2019 Speech Banking System

    19/24

    Speech Recoitio Techi6ue&

    !roadly speaking there are three approaches to speech recognition namely CED

    Acoustic .honetic approach%

    .attern ecognition approach%

    Artificial Intelligence%

    RAM/ 0102 is a memory &hich &ill store the S#eech Reco )n Ram Location

  • 8/12/2019 Speech Banking System

    20/24

    Bibliography

    PA#E NO : 78

  • 8/12/2019 Speech Banking System

    21/24

    TITLE AUT9OR

    9undamentals of Micro R processors ! % am

    and Micro R computers

    Micro R processor Architecture .rogramming % 7aonkar

    Applications

    .rinted 5ircuit !oard $esign 0echnology & % !ossart

    Micro R processor $ata "and !ook ;;;;;;

    !.! .ublications

    Micro R processor : !orole

  • 8/12/2019 Speech Banking System

    22/24

  • 8/12/2019 Speech Banking System

    23/24

    C:D abiner' *9undamentals of speech recognition,' >F;>E' chp B in 0he Speech Signal' .earson

    4dition' BFF>%

    CBD 3%6 9langan',Speech Analysis' Synthesis and .erception' Bnd ed% Springer;-erlag' #8'

    :LQB%

    CED abiner' *9undamentals of speech recognition,' Q@;@:' 5hp B in 0he speech signal' .earson

    4dition' BFF>%

  • 8/12/2019 Speech Banking System

    24/24