Welcome to NeuroWeb, International Neural Machine's (INM) WWW home!
INM applies neural network technology in development of pattern recognition
products for the software and technology marketplace. Such applications include
award-winning custom and shrink-wrapped optical character recognition (OCR/ICR)
products, and speech synthesis and decision support systems for variety of vertical
markets.
http://www.ineural.com/home.html
Optical Character Recognition ... OCR:
As the Internet becomes more and more widely used, much of the conventional paper
documents will be converted to electronic format and stored on-line. OCR eliminates
the time, error and expense of rekeying this data. INM's NeuroTalker OCR provides
you the means of converting these documents with proven reliability and accuracy.
The NeuroTalker OCR line currently consists of two products; OCR (Standard Edition)
and OCR AVE (Audio Visual Edition) for the visually impaired.
The NeuroTalker OCR:
NeuroTalker OCR offers the best price-performance in the industry today! This is the
lowest-priced, feature-reach product on the market! Only the most rudimentry,
no-frills OCRs are offered at this price level by our competitors.
Languages Support: Currently there is a built in dictionary for English.
Development is proceeding on support for all major European languages and
will be announced in the near future.
Typeface Independence: NeuroTalker OCR recognizes all commonly used fonts.
Fonts can be mixed on a page, paragraph or line.
Performance: More accurate on recognizing numerical data, technical terms,
names and any other text which cannot be easily corrected through an
embedded speller. Most conventional OCR's suffer serious performance drops on
such data, where as NeuroTalker OCR does not! NeuroTalker OCR delivers fast,
efficient and accurate text preservation. Now that is what OCR is all about!
Text/Graphics Separation: Our software incorporates advanced
text-from-graphics separation. In many cases it allows for efficient, fast and
accurate processing of pages without framing the text areas. Any frames
designated as graphics are automatically extracted and stored as separate
bitmaps together with the corresponding text files.
Built-in Spell Checker: Errors can be marked or auto-corrected. Proofing is
optional and can be enabled by the user at any time. While NeuroTalker OCR
comes with a master dictionary containing over 100,000 words, you can also
build your own custom dictionaries for special words and use them with the
click of a button.
Document Proofing:Your proofing time will be reduced using NeuroTalker OCR's
integrated Proof Reader. By displaying the actual portion of the bitmap
surrounding the suspicious characters you can immediately correct possible
errors by choosing from a list of alternate selections.
Application Compatibility: Once your text has been recognized you can import it
into your favorite Word Processor. NeuroTalker OCR maintains all layout
information and allows you to save documents in ASCII, Smart ASCII and RTF
formats.
Scanner Support: NeuroTalker OCR works with all TWAIN-compliant scanners. (
Hewlett-Packard, Fujitsu, Epson, Microtek and many others. )
System Requirements.486DX/Pentium with minimum 8 Meg RAM. MS
Windows 3.x, Windows 95, Windows NT.
User Support: Should you encounter any problems INM offers a variety of
support options including WWW and a dedicated toll free support line.
Neuro Talker Demonstration Download Site:
http://www.ineural.com/usr_info.html
NeuroTalker OCR AVE.
All In One
Custom Adaptive Applications
The NeuroTalker OCR AVE is a special edition of INM's OCR product line which provides both audio and
visual enhancements to a standard text menu interface, turning a PC into a reading
machine. While OCR AVE can be used as a stand-alone product, it will also integrate
with a large variety of existing access solutions giving it greater flexibility. OCR AVE
converts page images into text which can then be used by a large variety of
applications, such as Microsoft Word or WordPerfect. A summary of features include:
Languages Support: Currently there is a built in dictionary for English.
Development is proceeding on support for all major European languages and
will be announced in the near future.
Typeface Independence: NeuroTalker OCR recognizes all commonly used fonts.
Fonts can be mixed on a page, paragraph or line.
Performance: More accurate on recognizing numerical data, technical terms,
names and any other text which cannot be easily corrected through an
embedded speller. Most conventional OCR's suffer serious performance drops on
such data, we do not!
Text/Graphics Separation: Our software incorporates advanced
text-from-graphics separation. In many cases it allows for efficient, fast and
accurate processing of pages without framing the text areas.
Adjustable font size and screen magnification.
Built-in Spell Checker: Errors are automatically identified and corrected to
mimimize the need for user interaction.
Output Support: Once your text has been recognized you can either have the
document read back to you using the integrated audio text browser and/or
export it to an ASCII file.
Scanner Support: NeuroTalker OCR AVE works over 42 different scanners.
(Those supported include the Hewlett Packard line.)
Speech Devices: OCR AVE supports any speech card which is able to receive text
speech from either a serial (COM) or parallel (LPT) port.
Adaptive Product Support: OCR AVE is designed to be a stand-alone product
without the need for additional enhancement products such as screen readers
and screen enlargers. It offers the user the ability to tailor the interface to their
needs regardless of what they may be! As well, you could choose to use your
existing adaptive products and easily configure OCR AVE to work with them.
Automated Features: OCR AVE will automatically detect and correct a page's
orientation, as well as preserve its reading order!
System Requirements: 486DX/Pentium with minimum 8 Meg RAM. DOS
version 5.0 or a DOS Compliant Window which supports DOS version 5.0 or
higher.
User Support: Should you encounter any problems INM offers a variety of
support options including WWW and a dedicated toll free support line.
Text to Speech ... TTS: Converting electronic text to speech requires the use of an engine which is able to
determine the correct phonetic sequence for each word. The phonetic sequence is
then translated into waveforms which may then be played on audio-enabled PCs and
sound cards. INM uses the NeuroTalker Engine to achieve high accuracy when
translating text into speech.
Using the phonetic translation INM is able to provide realistic sounding speech with
natural intonation and stress. Currently INM's TTS engine supports American English
with development plans for major European languages. The engine also allows the
user the ability to add specialized pronunciation words and rules to the speech
database.
NeuroTalker TTS adheres to Microsoft's speech API definitions and can be integrated
into any application following these specifications. For more information on these
specifications contact Microsoft or refer to the MS Speech SDK.
Neuro Talker Products:
http://www.ineural.com/products.html#ocr_standard
Email: SDK@ineural.com
SUGGESTION: When testing ANY demonstration software in MS Windows ... it is suggested that:
1. Use an "Independent" Uninstall program/utility.
[Even when uninstall is included in the test application].
2. Backup copies of win.ini, system.ini, progman.ini to be held off-disk and "on floppy".
3. A zipped copy of the demonstration software should be held off-disk for backup.
GOOD HUNTING AND ENJOY!
SUPERADAPTOID.