Speech Recognition Essay

Only available on StudyMode
  • Download(s) : 220
  • Published : April 13, 2011
Open Document
Text Preview
Speech recognition (also known as automatic speech recognition or computer speech recognition) converts spoken words to text. The term "voice recognition" is sometimes used to refer to recognition systems that must be trained to a particular speaker—as is the case for most desktop recognition software. Recognizing the speaker can simplify the task of translating speech. Speech recognition is a broader solution which refers to technology that can recognize speech without being targeted at single speaker—such as acall system that can recognize arbitrary voices.

Speech recognition applications include voice user interfaces such as voice dialing (e.g., "Call home"), call routing (e.g., "I would like to make a collect call"), domotic appliance control, search (e.g., find a podcast where particular words were spoken), simple data entry (e.g., entering a credit card number), preparation of structured documents (e.g., a radiology report), speech-to-text processing (e.g., word processors or emails), andaircraft (usually termed Direct Voice Input).

PROJECTORS

Opening Note: When buying a home cinema projector, it is better to go for a true 16:9 widescreen projector rather than a 4:3. This is because a 4:3 screen has a lot of pixel wastage projecting black lines top and bottom. If you want an LCD home cinema, buy a Panasonic PTAE500. If you want a DLP one, get an Optoma H56 (or H76 if you can afford it). 

Four Main things to look for: 
1). The type of projector - LCD, DLP, CRT 
2). The Lumens output (brightness) of lamp - 800-1000 is good for LCD, DLP at 1000 is good. Dont get it to high or colours will look poor, less than 1500 is really needed for good colour reproduction.  3). The Resolution. For optimal picture make sure to get XGA or WXGA. A cheaper method is VGA/SVGA, but if you can get XGA or higher you wont regret the moeny spent.  4). A very key point is contrast ration. 300:1 and less is poor meaning blacks look dark grey. 700:1 and higher is good giving darker blacks and whiter whites (contrast is difference between these two). 1000:1 + is excellent. 

DLP Projection 

DLP technology is based on an optical semiconductor called a DMD chip (Digital Micromirror Device), which was invented in 1987 by Texas Instruments (cant remember the guys name). The way it works is that a DMD chip is made up of millions of tiny tiny mirrors, that can rotate at a 10 degrees angle (12 degrees on better models). These mirrors are literally capable of switching on and off thousands of times per second and are used to direct light towards and away from a dedicated pixel space. The lenght of time a pixel is on/off determines the level of gray seen in the pixel. Completely on is lightest (white) and off is darkest (black), although neither are ever pure white or black. The current DMD chips can produce 1024 shades of grey. However, colour is still to be added.  DLP Projection: (note its an old image thus only 3 colour segments) 

[pic] 

On a DMD chip, colour is added using a colour wheel. This is a wheel that spins round anything from 150x per second to 250x and more. Most of the cheaper models will be made up of a 4 segment colour wheel. This will be Red/Green/Blue and usually a clear segment. As light is passed through a point on the spinning colour wheel the mirrors switch in accordance to the light. 4 segment colour wheels can produce an effect that is known as rainbow. This is when the projector, in effect, struggles to change between all the colours quick enough, and along a line, the viewer may see a rainbow affect (See fig 1): 

[pic] 

6 Segment colour wheels are not effected in this way as badly as 4 segment colour wheels. Note that most viewers will not notice any rainbow affects, but some may. 6 Segment Colour wheels tend to spin faster and are made up of 2 sets of Red/Green/Blue segments thus enabling faster changes between the colours. Some projectors like theMarantz system (Absolutely amazing projector) have...
tracking img