lab 2

docx

School

CUNY College of Staten Island *

*We aren’t endorsed by this school

Course

117

Subject

Computer Science

Date

Jan 9, 2024

Type

docx

Pages

4

Uploaded by nerdinebenamor123

Report
CSC 117 Introduction to Computer Technology Lab #9 Voice Recognition Objective
1. The objective of this lab is to clearly understand the uniqueness of voice recognition on a computer. Description of the lab procedure 2. To start this lab you will need to do some research on voice recognition and answer the following questions: Which sectors of society use speech recognition programs?How accurate (in terms of percentage of words recognized) are current speech recognition programs? And What can go wrong with speech recognition? “Automatic Speech Recognition (ASR) is the technology that allows the computer to recognize (and even understand) the words that a person speaks into an electronic device”. To get started in the lab you will want to make sure that your headphones are correctly attached to the computer. The mike is input and the speaker is output. Now open Windows Speech Recognition by searching in the search bar located at the bottom left of the screen. Select Headset with Microphone and click NEXT. The program will ask you to read a few texts alou to make sure that the system can hear you. Continue with the prompts shown on the screen. If you then see a screen asking you to improve speech recognition accuracy you will have to select the ‘Disable document review’. As you continue through the appropriate steps you may be asked to select an activation Mode: you will just select ‘use manual activation mode’ and then click NEXT. You will then have to unlock the checkbox next to ‘Run Speech Recognition at startup’ on the following window. Next, you will click Skip Tutorial to move onto the next step to be able to finish up the setup. You will now see a Speech Recognition Microphone Widget floating on the top of the screen on your desktops Home Screen. Now you will have an open user boc. Click the new box, and type in your name. Keep all other defaults. Click next and follow the New User as the volume, audio quality and different prompts. For selection for dictation you will choose the first choice. Make sure not to have your profile updated with your documents or email. When asked for how to begin, you will start the tutorial and go through the first few lessons that way you get familiar with the program. Now you will practice without any prompts. Open a new WORD document and speak a few paragraphs from any link or thing you wish to read off of. Now
for the experimentation part of the lab you will learn more about the speech recognition systems. To calculate the Word Error Rate (WER) for evaluation of the speech recognition systems you must follow the following steps. S + D + I = N. S is the number of substitutions of incorrect words instead of correct words. D is the number of words that were left out. I is the number of extra words added by the speech recognition program. And N is the total number of words in the correct section that was detected. Now that you understand this equation you will put it to the test. First make sure that your speech recognition program and an empty WORD document are open. Now read a paragraph of at least 100 words from any website. “To start a new paragraph, say “new paragraph”, for a comma, you say, “comma”, and for a period, say “period”, and for a colon say “colon”. Do not incorporate formatting or quotes for this experiment.” Once you have read your paragraph, answer the following statements: Total number of words in the selection, total number of errors, words that were incorrect, and word error percentage. Repeat this exact same process for 3 other paragraphs. Again make sure a new WORD document is open before you start. 3. Observations/ Code What are the purposes of a speech recognition program? The purpose of the speech recognition program is to convert spoken words into text and allow someone to type handsfree. This is very helpful when it comes to people with disabilities who aren't able to type on a keyboard. What is the metric with which we can evaluate speech recognition programs? The metric is + + / N. 𝑆 𝐷 𝐼 S is the number of substitutions of incorrect words instead of correct words. D is the number of words that were left out. I is the number of extra words added by the speech recognition program. And N is the total number of words in the correct section that was detected. What parameters are there in the testing of speech recognition programs? Some parameters consist of the accuracy of the speech recognition, the ability for the speech recognition to pick up words with background noise, and how quickly the speech recognition can keep up on words without any error.
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
5. References 1. Lab #9 Voice Recognition