Добавил:
Upload Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Позааудиторне читання 2 курс КН.doc
Скачиваний:
1
Добавлен:
05.09.2019
Размер:
1.34 Mб
Скачать

Terminals

A terminal consists of a keyboard and a screen so it can be considered an input device, especially some of the specialized types. Some come as single units.

Terminals are also called:

  • Display Terminals

  • Video Display Terminals or VDT

A dumb terminal has no ability to process or store data. It is linked to minicomputer, mainframe, or super computer. The keyboard and viewing screen may be a single piece of equipment.

An intelligent, smart, or programmable terminal can process or store on its own, at least to a limited extent. PCs can be used as smart terminals.

A point-of-sale terminal (POS) is an example of a special purpose terminal. These have replaced the old cash registers in nearly all retail stores. They can update inventory while calculating the sale. They often have special purpose keys. For example, McDonalds has separate touchpads for each food item available.

Multimedia input

Multimedia is a combination of sound and images with text and graphics. This would include movies, animations, music, people talking, sound effects like the roar of a crowd and smashing glass.

Sound Input

Recording sounds for your computer requires special equipment. Microphones can capture sounds from the air which is good for sound effects or voices. For music the best results come from using a musical instrument that is connected directly to the computer. Software can combine music recorded at different times. You could be a music group all by yourself -singing and playing all the parts!

Voice Input

Voice input systems are now becoming available at the local retail level. You must be careful to get the right system or you'll be very disappointed.

Decide first what you want to do since a voice input program may not do all of these:

Data entry  - Talking data into the computer when your hands and eyes are busy should certainly be more efficient. You'd have to be very careful about your pronunciation!

Command and control - Telling the computer what to do instead of typing commands, like saying "Save file". Be careful here, too. The dictionary of understood words does not include some of the more "forceful" ones.

Speaker recognition - Security measures can require you to speak a special phrase. The computer must recognize your voice to let you in.

Speech to text - Translating spoken words direct to type would suit some authors just fine. You'd have to watch out for those "difficult to translate" phrases like "hmmm" and "ah, well, ... ummm."

(Nov. 2000) - Speech recognition and speech-to-text software is getting better rapidly. At least one retail version has the training period down to 5 minutes and ignores those filler phrases like "ummm".  This is getting exciting and actually useful!

A number of companies are now using speech recognition in their telephone systems. For example to find out what your bank account balance is, instead of punching in your account number on the phone keypad and choosing option 3 for current balance, you could speak your account number and say "Current balance". The computer will even talk back and tell you what it thinks you said so you can make corrections.

How do they change voice to data??

1. Convert voice sound waves to digital form (digital signal processing -DSP)

2. Compare digitized voice input to stored templates

3. Check grammar rules to figure out words

4. Present unrecognized words for user to identify

Types of Voice Recognition systems

Speaker dependent system

The software must be trained to recognize each word by each individual user. This might take hours of talking the dictionary into the computer, to be optimistic.

Speaker independent system

The software recognizes words from most speakers with no training. It uses templates. A strong accent would defeat the system, however.

Discrete speech recognition

The speaker must pause between words for the computer to tell when a word stops.

Continuous speech recognition

The speaker may use normal conversational flow.

Natural language

The speaker could say to the computer "How soon can we ship a dozen of product #25 in blue to Nashville?" - and get an answer!! Science fiction has come to life!!!