Scribblingpad

BSNL Shuns Huawei

2005-12-10T08:36:00.000-08:00

Indian state-run operator Bharat Sanchar Nigam Ltd. (BSNL) has yet to make a final decision on how long to cut off Huawei Technologies Co. Ltd. after it defaulted on a contract, but it appears the Chinese equipment vendor is already out in the cold when it comes to future business.

BSNL issued a notice to the vendor in October, asking for an explanation as to why it shouldn't be blacklisted from future contracts after its local partners failed to deliver 1.05 million lines of CDMA2000 equipment. (See BSNL Threatens to Ban Huawei.)

BSNL has been agonising over a decision for weeks. According to Huawei spokeswoman Cassandra Cheong, "Based on our understanding, BSNL has yet to make a decision and we are still in discussions to continue our long, successful relationship together."

But in the meantime BSNL has cancelled the CDMA deal and is due to put it back out for tender, excluding Huawei and its partners of course.

The episode has delayed its rollout of much-needed phone lines in rural India -- there's a waiting list of around 1.35 million -- and it reportedly asked ZTE Corp. (Shenzhen: 000063 - message board; Hong Kong: 0763) to take on the contract to help out. ZTE submitted the second lowest bid after Huawei in the original tender, but the rival Chinese vendor turned it down, saying it couldn't match Huawei's price.

With Huawei out of favor, China's other equipment vendors -- also known for their low prices -- are picking up extra business from the rapidly-expanding carrier.

India's Business Standard reports that Huawei has lost out to ZTE on a 50:50 joint venture with BSNL said to be announced soon. Both vendors had been shortlisted for the tender, but the carrier ultimately chose ZTE for the project that will include the manufacture of one million CDMA mobile handsets and one million fixed wireless terminals.

Neither BSNL nor ZTE had responded to requests for comment as this article was published.

The news follows an announcement from UTStarcom Inc. (Nasdaq: UTSI - message board) that it has been awarded an expansion contract from BSNL for 90,000 DSL lines, IP DSLAMs being another Huawei specialty. (See UTStarcom, BSNL Team and Huawei Deepens DSLAM Penetration.)

Huawei's taking it bravely and is optimistic it'll pick up more business, having worked with BSNL since 1999. "We are unable to comment on other companies' contract[s], but Huawei remains committed to the Indian market," Cheong writes in an email to Light Reading. "We believe Huawei will be able to positively contribute to India's development by generating additional opportunities for the India telecom market."

The ongoing quarrel with partly-state-owned BSNL aggravates an already delicate situation for Huawei, which is viewed with suspicion by the Indian government and has had its expansion plans delayed repeatedly. (See India Blocks Foreign Telecom Gear.)

As for the proposed ban, the Indian media is reporting that while senior officials at BSNL want to cut Huawei off completely, there's a consensus at the board level that Huawei should be blacklisted from all future CDMA tenders, but still be allowed to bid on its upcoming mega GSM tender for 60-70 million lines, worth around $4.5 billion.

The delay in making the decision is mostly about money; BSNL knows that Huawei is likely to quote much lower prices for contracts than its North American and European competitors and even its Chinese counterparts. And with India's telecom market growing rapidly, BSNL needs to build out a large amount of infrastructure to meet demand and wants to do it as cheaply as possible.

proxy sites ----circumventor

2005-11-01T01:01:00.000-08:00

http://www.peacefire.org/circumventor/simple-circumventor-instructions.html
http://www.peacefire.org/circumventor/simple-circumventor-instructions.html

softwares for remote desktop access

2005-10-30T14:27:00.000-08:00

http://www.downloadjunction.com/product/software/5723/index.html

place for cracks

2005-10-30T14:11:00.000-08:00

http://cracks.thebugs.us/last.shtml?186670

Email change confirmation required

2005-10-23T11:03:00.000-07:00

Hello srinivas,

You or someone with the IP address 220.225.137.242 have requested to change the email address associated with your fileUPYOURS.com account.

To confirm the change of your email address click on the link below. Otherwise discard this message and your email address will not be changed.

http://www.fileupyours.com/account.php?action=confirm_email_change&userid=3484&code=XbDieHo5rQFmi0Faddes1aA5mpnqqjBdk4moBAjThP5qGBrJFv

--
Regards,
fileUPYOURS.com

speaking pen

2005-10-22T10:26:00.000-07:00

how to password protect a web page

2005-10-17T20:55:00.000-07:00

source: http://www.ehow.com/how_6226_password-protect-web.html

Password protection keeps only the users you want on your page. If you know JavaScript or even plain old HTML, you can use password protection.

Placing a Link from an Unprotected Page to a Protected Page

	Steps:
1.	Open the unprotected page in a basic text editor. Place the cursor where the link will appear.

2.	Type the following, substituting your own text or graphic: Click here to access the password protected page.


	Tips:
	Users must click on a link to access a page unless the first page of a site is protected. If the main page is the password protected page, start at Section 2.

	For the purposes of this section, the destination of the link, a script file, is called "password.html".


Creating a Script for the Password Page

	Steps:
1.	Open a basic text editing program. Begin by typing the opening tags: <html><head>

2.	Type the title of your page and close the title with a tag: <title>Protected Page: Please enter password</title>

3.	Type the following: <script language="Javascript"> This command tells the browser a script is about to begin.

4.	Press the Enter key and type the following: //prompt. Press the Enter key again.

5.	Enter the prompt by typing the following: var password = prompt("Enter in the password"""); There are three double-quotation marks after the word "password" and a semicolon after the last parentheses.

6.	Type the following, replacing "pass" with the word or numbers the password will be: if (password == "pass")

7.	Now type the following, replacing "correct.html" with the name of the page the correct user is directed to: {location = "correct.html"} This tells the browser to send the user who enters the correct password to this page.

8.	Send users who type the incorrect password to an error page by typing the following, replacing "error.html" with the error page you have created: else {location = " error.html"}

9.	End the script by typing the following: //--></script>

10.	Close the scripted section of the page by typing the follow: </head>.

11.	Add a body section for browsers that do not read JavaScript: <body>This page is designed for viewing by Netscape 2.0, Internet Explorer 3.0 or above. Please upgrade your browser and return soon!</body>

12.	Close the Web page by typing: </html> Save the page (for this example, save it as "password.html").


	Tips:
	Users who enter the correct password are taken straight to the page; users who enter the incorrect password are sent to an error page. You create the error page, called " error.html" here, in the next section.

	Be sure to enter all commands as shown. JavaScript will not work if the command isn't exactly right.


Creating the Error Page

	Steps:
1.	Open a basic text editing program. Begin by typing the opening tags: <html><head><title>

2.	Title the page "Error" or "Wrong Password": <title>Sorry ' Wrong Password!</title></head>

3.	Type the error message the viewer will see, such as: <body> <h1>I'm sorry, the password you typed is incorrect.

4.	Give the user an opportunity to re-enter the password: Please click here to try again. This allows the user the opportunity to return to the password script.

5.	Allow the user to back out: <p>Otherwise, click here to return to the main page. </h1>("here.html" indicates the name of the page that links to the protected page)

6.	Close the page by typing the closing tags: </body></html>


	Tips:
	Format this page to resemble your other pages, so viewers don't get lost.


Testing the Password Script

	Steps:
1.	Open a Web browser. Select Open from the File menu, locate the page that links to the password-protected page, and click Open to view it.

2.	Click on the link to the password-protected page.

3.	You will be prompted for the password. Make sure you enter it correctly; passwords are case-sensitive.

4.	Gauge the results. If it does not work, go back and double-check that you entered the script correctly.


	Tips:
	Try typing both the correct and incorrect passwords so you will see what your visitors see.


	Warnings:
	Do not link back to the password file (password.html). If the person does not know the password, they will be endlessly looped and cannot get out of the script.

Finding More Complex Password-Protection Scripts

	Steps:
1.	Visit pages that offer free JavaScripts and do a search for "password". Many of these scripts are just cut and paste, with minimal changes.

2.	Protect extremely sensitive information by hiring a Web developer (or your ISP) to create a customized password script for you. These scripts are extremely sensitive and very difficult for hackers.

3.	Contact your Internet Service Provider and ask if they provide password protection services. This entails setting aside a portion of the server for your pages, so some ISPs may charge extra.

4.	If you use a free script, add a courtesy button or reference to the provider on your page.

5.	You may need to get permission before using JavaScripts from another Web site. When in doubt, ASK!

6.	Not all browsers support scripts. Be sure to follow directions for providing alternates to scripts on your Web page.

different ways of writing an ip address

2005-10-17T20:38:00.000-07:00

Here the information concerning converting to IP address to a number string:
>
> With some firewall implementations, you can bypass the URL filtering rules:
>
> Obtain IP address of a filtered URL: http://www.netaddress.com or
> http://204.68.24.100
> Convert IP address to binary using calc.exe - 11001100 01000100 00011000
> 01100100
> Concatenate the binary together 11001100010001000001100001100100
> Convert this number to decimal using calc.exe - 3427014756
> Enter this in your browser - http://3427014756

100 + 24*256 + 68*256^2 + 204*256^3 = 3427014756

Is a lot less work than doing the middle conversion.

Actually, (for those who think in reverse-Polish notation, this
is the obvious way), the easiest is more like,

((((204*256) + 68)*256 + 24)*256 + 100 = 3427014756

On calc.exe, keystrokes are like,

204*256+68*256+24*256+100

Of course, another fun way to get by stoopid filters that don't really
understand URLs is to,

% printf "0%o.0%o.0%o.0%o\n" 204 68 24 100
0314.0104.030.0144

To give,

http://0314.0104.030.0144/

Or even just convert 3427014756 to octal,

http://031421014144/

Or how about mix and match,

http://204.0104.24.0144/

Or,

http://204.0104.6244/
http://204.0104.014144/
.
.
.

There are a _lot_ of ways to write an IP address in a URL.
--
Crist J. Clark Network Security Engineer
crist.clark@globalstar.com Globalstar, L.P.
(408) 933-4387 FAX: (408) 933-4926

source...:http://lists.jammed.com/pen-test/2001/04/0097.html

Account activation

2005-10-09T05:38:00.000-07:00

Hello srinivas,

Thank you for registering. fileUPYOURS.com requires that your account be activated. To do so, click on the link below. Your account will not be accessible until it is activated.

http://www.fileupyours.com/account.php?action=activate&userid=3484&code=jeLuAoWE1V

If you did not register for an account at the domain above, please discard this email message, we apologize for any inconveniences.

--
Regards,
fileUPYOURS.com

Some very useful websites

2005-10-09T05:21:00.000-07:00

Hostings Without ads

.http://www.liquid2k.com/(recommended)
.http://www.brinkster.com/ (recommended)
.http://www.demented.org/
.http://www.cfm-resources.com/
.http://www.digitalrice.com/
.www.free.fr
.Shyper.com
.http://www.freewebtown.com/community/index.php
.http://www.ocean9.com
.http://www.anewbay.com
.www.x10hosting.com
.x10hosring.com has 2.5 gigs free
.http://www.hostingzero.com/hosting.php
.www.t35.net

Host support PHP, 15-25 MB, MySQL, Some Ads

.http://www.host.sk/ (no-ads)
.http://www.php50.com/ (pop-up)
.http://www.t35.com/ (pop-up)
.http://www.clawz.com/ (banner)
.http://www.tripod.co.uk/ (pop-up, MySQL included)

Large Host for files and FTP

.http://www.freewebsites.com/
.http://www.gizba.com/
.http://www.webavenue.org/
.http://www.hut.ru/
.http://www.web1000.com/
.http://www.phidji.com/ (100MB, no ads, ASP)
.http://www.spacetowns.com ( 50 MB, no ad's )
.http://www.freecoolpages.com ( 50 MB, PHP4 )
.http://www.topcities.com ( 150MB )
.http://www.nerocities.com ( 100MB, FTP, no ad's )
.http://www.Topzite.com ( 1000MB, FTP, no ad's, CGI )
.http://www.tripod.co.uk ( 100MB, ad's )
.http://www.mystigall.com/ (75MB, no ads, FTP, CGI)
.http://www.webazn.net/ (50MB, no Ads)

Host Some ads, small MB

.http://www.enjetek.com ( 15MB )
.http://www.f2g.net ( 15MB, FTP, no ads, PHP4 )
.http://www.geocities.com/
.http://www.angelfire.com/
.http://www.tripod.com/
.http://www.topcities.com/

Host Support CGI for Ikonboard and YaBB, FTP

.http://www.brinkster.com/ (no FTP, and no ads)
.http://www.raketnet.nl/
.http://www.cfm-resources.com/
.http://www.aspfreeserver.com/
.http://www.websamba.com/

Host >15MB, transfer 1GB/month, PHP, MySQL, CGI

.http://www.port5.com ( 15MB, FTP, no ad's )
.http://www.amzweb.net/signup.php (also offers ASP)
.http://www.spaceports.com/
.http://www.digitalrice.com/ (only 5 MB space)
.http://www.portland.co.uk/ (only 100 MB transfer)

Host 20-30MB, PHP, MySQL, FTP

.http://www.host.sk/ (no ads)
.http://www.spaceports.com/
.http://www.tripod.co.uk/
.http://www.multimania.com/
.http://www.amzweb.net/signup.php
.http://www.web1000.com/ (use at your own risk! vious2)
.http://youthofamerica.org/ (20MB, browser noads RealVideo RealAudio POP Email ASP SSI CGI-BIN)
.http://www.t35.com ( 35MB, FTP, no ad's )

Host > 20MB , transfer unlimited, FTP

.http://www.webavenue.org/ (no ads)
.http://www.ghs20.com/ (no ads)
.http://www.host.sk/ (no ads)
.http://www.amzweb.net/signup.php
.http://www.netfirms.com/
.http://www.fateback.com/
.http://www.barrysworld.com/ (35MB, no ads, fast, PHP)

~~~Free BulletinBoards/Forums

.http://www.phpbb.com - phpBB2 (PHP, mySQL)
.http://woltlab.de - wbb1 (PHP, mySQL)
.http://yabb.info - YaBB SE (PHP, mySQL)
.http://openbb.com - OpenBB2 (PHP,mySQL)
.http://www.invisionboard.com - Ibforums (PHP,mySQL AND CGI)
.http://forum.snitz.com Snitz Forum2000 (ASP)
.http://german-bulletin-board.de - gBB (PHP, mySQL) (Cool)
.http://www.phorum.org/ - Phorum (PHP + MySQL; open source)
.http://www.xmbforum.com/ - XMB (PHP + MySQL)
.http://www.minibb.net/ - MiniBB (PHP + MySQL)
.http://www.yabb.com/- YaBB original (CGI)
.http://www.mybboard.com/ - MyBB/DevBB (PHP, MySQL)
.http://www.xmbforum.com/ - XMB (PHP, mySQL
.http://www.simplemessageboard.com/ - SMP (CFM, MSsql)
.http://www.lokwa.com/ - lokwaBB (PHP, Mysql)
.http://www.mercuryboard.com/ - MurcuryBoard (PHP, MySQL)
.http://edge-programming.com/eboard/ - EBoard (PHP)
.http://www.smartbb.net/ - SmartBB (PHP, Mysql)
.http://www.icredibb.com/ - iBB (PHP, Mysql)
.http://www.myboard.co.uk/ - myBoard (PHP, MySQL)
.http://mybboard.com - DevBB (PHP, mysql)
.http://www.versiforum.com - VersiForum (ASP)

~~~Free Remotly Hosted Forums

.http://upperboard.com/
.http://forums.asp-dev.co.uk/
.http://www.xsorbit.com/
.http://www.proboards.com/
.http://www.ezboard.com/
.http://forums.hostultra.com/
.http://www.network54.com/
.http://www.everyone.net/
.http://www.mycool.com
.http://www.guestforum.com/
.http://www.voy.com/
.http://www.dk3.com/
.http://www.homepagetools.com/ultraboard
.http://www.swiftsolution.com/
.http://groups.yahoo.com/
.http://communities.msn.com/
.http://on.starblvd.net/meet
.http://forums.delphi.com/
.http://www.suddenlaunch.com/
.http://bb.bbboy.net/
.http://www.eboards4all.com/
.http://www.bizhat.com/
rapidforums.de

~~~Free Subdomains

.www.freeurl.com
.www.jwdx.com
.www.cjb.net
.www.dot.tk
.www.v3.com
.www.hotredirect.com
.www.internetjump.com
.www.explode.to
.www.zdos.com
.www.webalias.com
.www.reduce.to
.www.warping.to
.www.gosurfto.com
.www.2000c.net
.www.shorturl.com
.www.has.it
.www.doze.to
.www.nigx.net
.www.1fx.net
.www.soar.to
.www.123redirect.com
.www.ipfox.com
.www.webweaver.nu
.www.ohgo.com
.www.url-redirection.org
.www.aliasnames.com
.www.get-2.com
.www.myredirector.com
.www.dot.nu
.www.tr.cx
.www.kickme.to
.www.rapworld.com/url/
.www.dk3.com
.www.xiy.net
.www.url.animeumbrella.com
.www.suite.net/url.htm
.www.surftohere.com
.www.israd.net
.www.ontheweb.nu
.www.globalredirect.com
.www.flash.to
.www.zooming.to
.www.emu.vg
.www.linkworld.to
.www.rename.net
.www.url4life.com
.www.n2v.net
.www.nethop.com
.www.webmask.com
.www.heroffice.com
.www.crcpl.tsx.org
.www.guruguru.to
.www.iscool.net
.www.dkanet.com
.www.r67.com
.www.uni.cc
.www.fr.fm
.www.d.bz
.www.e33.de
.www.de.vu
.www.dd.vu
.www.6x.to
.www.b4.to
.www.b6.to
.www.h3.to
.www.thx.to
.www.rox.to
.www.faster.as
.www.hop2.de
.www.tsx.to
.www.tsx.org
.www.bootme.to
.www.2fbi.de
.www.ubb.cc
.www.IDz.net
.http://zwap.to
.www.xsub.ws
.www.2b.to
.www.elite.to
.www.2fbi.de
.www.suckz.de
.www.b6.to
.www.quickurl.com
.http://www.9cy.net/myred/signup.php

~~~Free Domains

.http://www.dot.tk/ (redirection)
.http://www.uni.cc/ (redirection, webhosting, OR domain)
.http://www.ripn.net:8080/en/ (domain)
.www.dhs.org
.http://www.acidmist.com/
.http://www.1millionfreedomains.com/ (requires reserve fee)
dndiscuss.com get a free .com .net .info for some posts

~~~Co-Brands

.http://megagiga.com
.http://portland.co.uk
.http://www.communityarchitect.com/
.http://www.muttle.com/

~~~Free FileStorage

.http://java.isavvix.com/freeback.jsp
.http://www.myplay.com/
.http://www.sharemation.com/
.http://www.staroffice.com/
.http://www.storagevault.net/
.http://www.tiomeg.com/
.http://briefcase.yahoo.com/
.http://www.zden.com/Signup.asp?ID=136463
.http://rapidshare.de/en/main.html
.http://up.li.ru/
.http://www.yousendit.com/
.http://www.walagata.com/
.http://www.photobucket.com/
.http://www.uploads.biz/
.http://www.solisstyle.com/ims/splash.php
.http://www.imagehosting.us.com/
.http://www.bestupload.com/
.http://www.hostmyimages.net/
.http://www.realimagehost.com/
.http://www.netpix.org/
.http://www.imgzone.info/
.http://host.pixerver.com/
.http://www.iuploads.com/
.http://www.hostmyfile.net/
.http://www.flypicture.com/
.http://upload.refrozen.com/
.http://www.picvault.info/
.http://www.filehosting.org/
.http://www.xs.to/
.http://www.uploadimages.net/
.http://www.hostmyfile.net/
.http://dinghy.homeip.net/upload.php
file types allowed: gif, jpg, jpeg, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe
file size limit: 737.28 MB
.http://zachhorn.com/.upload/index.php
file types allowed: gif, jpg, jpeg, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe
file size limit: 102.4 M
.http://psplus.tk/fileupload/index.php
file types allowed: mp3, wma
file size limit: 100 MB
.http://members.lycos.co.uk/signifer123/
file types allowed: gif, jpg, jpeg, png, fla, swf, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe
file size limit: 100 MB
.http://www.hellreapers.com/fileupload/
file types allowed: gif, jpg, jpeg, png, zip, rar, w3g, wmv
file size limit: 58.828 MB
.http://www.al-nahda.com/kutub/
file types allowed: gif, jpg, jpeg, png, txt, nfo, doc, pdf, rtf, htm, dmg, zip, rar, gz, tif, tiff
file size limit: 49.063 M
.http://www.objext.com/upload/
file size limit: 20.006 MB
.http://dhalkar.serveftp.com/fileupload/
file types allowed: png, css, bmp, gif, jpg, jpeg, png, txt, nfo, doc, rtf, dmg, zip, rar, gz
file size limit: 20 MB
.http://www.b3n.cc/filehost/
file types allowed: gif, jpg, jpeg, png, txt, zip, rar, gz, exe
file size limit: 14.648 MB
.http://www.developingwebs.net/submissions/fileupload/index.php
file types allowed: gif, GIF, jpg, JPG, jpeg, png, txt, doc, htm, html, zip, rar, bmp, tif
file size limit: 11.819 MB
.http://www.da_insane.boo.pl/d/upload/index.php
file types allowed: gif, JPG, jpg, jpeg, JPEG, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe
file size limit: 10 MB
.http://www.nitroushosting.net/boogiedown/fileupload/
file types allowed: psd, gif, jpg, jpeg, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe
file size limit: 9.866 MB
.http://boffy.3b1.org/upload.php
file types allowed: gif, jpg, jpeg, jiff, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, tar, bz2, gz, exe
file size limit: 5 MB
.http://www.graphics-tech.com/mural/
file types allowed: gif, jpg, jpeg, png
file size limit: 4 MB
.http://dev.dcemulation.com/bormod/upload/
file types allowed: gif, jpg, jpeg, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe
file size limit: 3.93 MB
.http://www.mydigitalevo.com/uploader/
file types allowed: gif, jpg, jpeg, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe, ppt, pub, doc
file size limit: 2.048 MB
.http://upload.3rd-wing.com/
file types allowed: mis, gif, jpg, jpeg, png, txt, nfo, doc, rtf, htm, dmg, zip, rar, gz, exe
file size limit: 2 MB
.http://212.33.90.248/~omega/upload/
file size limit: 2 MB
.www.69mb.com
.www.35mb.com
.www.streamload.com
.www.dropload.com
.www.omnilect.com
.http://free.hosthongkong.net/
.http://www.powow.com/
.http://www.filesupload.com/
.http://fastuploads.com/
.www.sendthisfile.com
.www.downloadhosting.com
.http://www.zippyvideos.com
.http://www.putfile.com
.http://www.mytempdir.com/
.http://www.webfile.ru
.http://www.updownloadserver.de/
no Traffic- or Downloadlimit
After Upload, you'll get an URL. This URL you can Post as an
Downloadlink or something else. Your file will be deleted, if after 30
days there where no download action detected
.http://www.datapickup.com/
Space: 1.5G
Limit: Unknown; None?
.http://uploadhut.com/
.http://www.4shared.com/

~~~Free Pop3Mails

.http://mail.uni.de
.http://quickemail.de
.http://gmx.net
.http://mypad.com
.http://kgb.cz
.http://amexmail.com
.http://freemail.web.de
.http://hotmail.com (not sure)
.http://uyeler.mynet.com
.http://eng.mail.port.ru
.http://passagen.se
.http://cjb.net
.http://myrealbox.com
.http://mail.urbia.de
.http://my-mail.ch
.http://westonline.com
.http://mundomail.net
.http://www.hotpop.com
.http://webmail.berlin.de
.http://club.lemonde.fr
.http://emailcity.de
.http://mail.lycos.co.uk
.http://epost.de
.http://schondrin.de
.http://freemail.nl
.http://idlo.de
.http://saintmail.net
.http://topsurf.com
.http://uk2.net
.http://liquid2k.com
.http://dostmail.com
.http://china.com
.http://hotpop.com
.http://newmail.net
.http://mail.arabia.com
.http://mail.ofir.dk
.http://home.se
.http://epost.portalen.no
.http://runbox.com
.http://mail.tut.by

~~~Free Email for your site

.http://bigmailbox.com
.http://zzn.com
.http://i-p.com
.http://oemmail.com
.www.gawab.net
.www.shinyfeet.com

~~~Free Counters/Stats for ur site

.www.sitemeter.com
.http://www.addfreestats.com
.www.bravenet.com
.www.admo.net
.www.thecounter.com
.http://www.extreme-dm.com/
.http://www.nedstat.com/
.www.stats4all.com
.www.realtracker.com
.http://www.analogstats.com
.http://www.geocounter.net
.www.webtrendslive.com
.http://www.okcounter.com
.http://www.vioclicks.com

~~~~Free image hosts

.http://www.photobucket.com
.http://www.imageshack.us
.http://www.daplayazclub.com/uploader/upload.php
.http://www.daplayazclub.com/uploads/
.http://www.movv.com/prvupload/upload.php
.http://www.movv.com/prvupload/uploads/
.http://reichnation.com/upload1.php
.http://www.reichnation.com/uploads/
.http://www.six-online.com/fileupload/
.http://www.six-online.com/fileupload/index.php
.http://www.pcmunkey.net - once they have there servers back.
.http://www.xbox-connection.com/hostedimages/index2.php
.http://www.xbox-connection.com/hostedimages/
.http://www.freepixhosting.com
.http://www.freepixhosting.com/uploadedfiles/
.http://www.9ximage.com/
.http://www.00gaming.com/00ImageHost/
.www.deanfarnham.net
.http://www.sighost.us
.www.photojerk.com

.http://fileupyours.com
Everything unlimited

.http://datapickup.com
Max File Size: 1.5GB
File deleted after 7 days

.http://freespace.filefront.com
Max File Size: 1GB (Files cannot be larger than 1GB or take more than 10 hours to upload)
Everything else is unlimited.

.http://www.yousendit.com
Max File Size: 1GB
File deleted after 7 days

.http://www.transferbigfiles.com
Max File Size: 1GB
File deleted after 5 days/20 downloads

.http://www.spread-it.com
Max File Size: 500MB
File deleted if not downloaded for 30 days

.http://ftpz.us
Max File Size: 498MB
Everything else is unlimited

.http://www.thefilehut.com
Max File Size: 250MB
Free registration required

.http://www.megaupload.com
Max File Size: 250MB (used to be 500MB)
File deleted if not downloaded for 30 days

.http://www.updownloadserver.com
Max File Size: 250MB (Uploads will be checked by an admin before publishing)

.http://www.sharebigfile.com
Max File Size: 250MB
File deleted after 7 days

.http://www.filecache.de
Max File Size: 125MB
File deleted if not downloaded for 30 days

.http://www.pushfile.net
Max File Size: 100MB
File deleted if not downloaded for 7 days

.http://www.dropload.com
Max File Size: 100MB
File deleted after 7 days or after the first download (yep.. it's true)

.http://www.turboupload.com
Max File Size: 70MB

.http://www.picapic.net
Max File Size: 60MB

.http://www.nexmicrosystems.com
Max File Size: 60MB
File deleted after 30 days

.http://www.savefile.com
Max File Size: 60MB
File deleted if not downloaded for 14 or 30 days. FAQ says 14, upload page says 30...

.http://www.psychohost.com
Max File Size: 50MB
Everything else unlimited.

.http://www.freeuploader.com
Max File Size: 50MB

.http://www.ushareit.com
Max File Size: 50MB

.http://sharefiles.ru
Max File Size: 50MB
File deleted after 30 days

.http://www.wirefiles.com
Max File Size: 50MB

.http://www.xshare.us
Max File Size: 50MB

.http://app02.bonpoo.com/file.htm
Max File Size: 50MB

.http://rapidshare.de
Max File size: 50MB
Download Limit: 50MB/hour
File deleted if not downloaded for 30 days

.http://qfile.de
Max File size: 50MB
Download Limit: 50MB/hour
File deleted if not downloaded for 30 days

.http://www.uploadhut.com
Max File Size: 50MB

.http://www.gigashare.com
Max File Size: 50MB (multiple files upload option)

.http://illhostit.com
Max File Size: 40MB

.http://www.filebuffer.net
Max File Size: 40MB
Everything else seems to be unlimited

.http://www.35mb.com
Max File Size: 35MB
Free registration required
Everything else is unlimited

.http://up.li.ru
Max File Size: 32MB
File deleted after 3 days (2 months if file extension is jpg/jpeg/png)

.http://sendmefile.com
Max File Size: 30MB
File deleted after 14 days

.http://www.swiftdesk.com
Max File Size: 30MB
Free registration required

.http://www.ultrashare.net
Max File Size: 30MB
File deleted if not downloaded for 30 days.

.http://falcon.o-wh.com
Max File Size: 30MB

.http://www.ripway.com
Max File Size: 30MB
Free registration required

.http://www.mytempdir.com
Max File Size: 25MB
File deleted after 14 days

.http://www.myfilestash.com
Max File Size: 25MB
Free registration required

.http://www.myfileshack.com
Max File Size: 25MB
Free registration required

.http://11mbit.de
Max File Size: 20MB

.http://www.webfile.ru
Max File Size: 20MB
File deleted after 7 days

.http://www.yourfilehost.com
Max File Size: 10MB
File deleted if not downloaded for 14 days

.http://www.da_insane.boo.pl/d/upload/
Max File Size: 10MB

.http://www.filesupload.com
Max File Size: 10MB
Free registration required

.http://www.4shared.com
Max File Size: 10MB
Free registration required

.http://come2store.com
Max File Size: 10MB
Free registration required

.http://webfilehost.com
Max File Size: 10MB
File deleted after 30 days

.http://www.filecabi.net/free-video-hosting.html
Max File Size: 10MB
File deleted if not viewed for 30 days

.http://www.putfile.com
Max File Size: 10MB
File deleted if not viewed for 6 months

.http://uppit.com
Max File Size: 6MB
File deleted if not downloaded for 30 days

.http://www.zippyvideos.com
Max File Size: 5MB
File deleted if not viewed for 60 days

.http://www.desearch.net
Max File Size: 4MB

.http://www.uploadx.net
Max File Size: 3MB

.http://www.filepost.us
Max File Size: 3MB
Free registration required

.http://www.thefilebucket.com
Max File Size: 2MB (not sure)
Free registration required

.http://storagefun.com
Max File Size: 2MB
Free registration required

.http://www.flypicture.com
Max File Size: 1MB
File deleted after 30 days

.http://www.bestupload.com
Max File Size: 600K
File deleted if not downloaded for 30 days

.http://www.fdrop.com
.http://www.exbyte.net
.http://filecache.de

.www.zupload.com
- unlimited download
- Resume Download : No
- upload filesize is 500 Mb

Plan from China's Huawei may be blocked in India---Report: Indian government cites security concerns

2005-10-01T10:32:00.000-07:00

AUGUST 17, 2005 (IDG NEWS SERVICE) - BANGALORE, India -- A request by China's Huawei Technologies Co. for a license that would allow it to bid on state telecommunications projects in India has met with opposition from certain sections of the government on security grounds, according to a report in The Times of India newspaper.

Huawei applied to the Indian government in March for a license that would allow it to bid on telecommunications projects from India's large, government-controlled telecommunications services companies, Gong Wenhe, director of business development at Huawei Technologies India Pvt., told the IDG News Service today.

According to the report in The Times of India yesterday, the government is evaluating the risks of exposing strategic Indian telecom networks to the Chinese because of fears that China could attack India's communications networks should relations between the countries deteriorate.

Indian government officials were unavailable to comment on the newspaper report.

Although China and India have grown closer politically in recent years, and Indian software outsourcing companies have set up operations in China, the two countries have a long-standing border dispute, which led to a war in 1962.

The license would allow Huawei's India subsidiary, Huawei Technologies India, to bid for installation and maintenance work, among other types of telecommunications projects in India. The subsidiary, based in Gurgaon, which is near Delhi, is a sales and marketing operation that sells equipment to service providers in the country.

"Our application is still pending," Gong said. He added that Huawei had not received any communication from the government about its apprehensions over security concerns. The Foreign Investment Promotion Board in Delhi, which processes foreign investment proposals, has met a number of times since Huawei submitted its proposal.

Huawei has sold equipment to private and government-controlled telecom service providers in the country. However, in deals involving large government-controlled service providers, such as Bharat Sanchar Nigam Ltd., the regulations require Huawei to partner with a local company if it wants to implement and maintain projects.

If the license request before the government is cleared, Huawei would be able to bid directly for telecommunications projects from government-controlled service providers, Gong said.

By doing project work directly, rather than working through implementation partners, Huawei would be able to cut its costs, offer better prices to customers and generate employment in the country, Gong said.

The company already has a research and development center in Bangalore that employs more than 900 people. Set up in 1999, the center was the first set up in India by a Chinese technology company.

Huawei has announced plans to invest $100 million in India. Its plans include expanding the R&D center and setting up a manufacturing facility in Bangalore that will be operational later this year, Gong said.

for a quick glance at interview questions in c and c++

2005-09-29T17:42:00.000-07:00

www.acetheinterview.com
www.moskalyuk.com
www.cpunivers.com

athi sookshma robo.....

2005-09-25T11:15:00.000-07:00

for java scripts, java applets, java codes etc....

2005-09-22T19:25:00.000-07:00

www.alienmenus.com
www.jars.com
www.freeapplets.com
www.applets.tripad.com
www.codeproject.com
www.javafile.com
www.codebrain.com

for java model papers

2005-09-22T19:24:00.000-07:00

www.jchq.net
www.java.about.com
www.javarach.com
www.valoxo.ch/jr/mocks

articles about huawei in frontline

2005-09-21T08:01:00.001-07:00

www.flonnet.com/fl2209/stories/20050506002500900.htm
www.flonnet.com/fl2206/stories/20050325003600400.htm

articles of huawei in frontline

2005-09-21T08:01:00.000-07:00

huawei...actual fear...!

2005-09-20T09:11:00.000-07:00

http://www.networkworld.com/news/2005/081705-huawei-india.html

huawei's and chineese govt....military

2005-09-16T08:01:00.000-07:00

http://www.techdirt.com/articles/20050819/1520213_F.shtml

about huawei

2005-09-15T20:27:00.000-07:00

http://voip-blog.tmcnet.com/blog/greg-galitzine/voip/up-up-and-huawei.html

Some useful websites for downloads

2005-09-09T12:17:00.000-07:00

www.download.com
www.freedownloadscenter.com
www.topshareware.com
www.hotdownloads.com
www.webattack.com
www.freewarefiles.com
www.simtel.com
www.freedownloads.be
www.jumbo.com
www.amazesoft.com
www.softhall.com
www.majorgeeks.com

some useful websites for resume building

2005-09-03T15:25:00.000-07:00

www.resumebuilders.net
www.aperfectresume.com
www.resume-place.com
www.careerbuilder.com
www.resume.monster.com
www.gresumes.com
www.e-resume.net
www.resumelogic.com
www.careerdesigns.com

http://computing.breinestorm.net/speaker+network+neural+identification+speakers/

2005-07-17T14:36:00.000-07:00

Paper: REUSABLE BINARY-PAIRED PARTITIONED NEURAL NETWORKS FOR TEXT-INDEPENDENT SPEAKER IDENTIFICATION :: Stephen A. Zahorian

REUSABLE BINARY-PAIRED PARTITIONED NEURAL NETWORKS FOR TEXT-INDEPENDENT SPEAKER IDENTIFICATION
Stephen A. Zahorian Department of Electrical and Computer Engineering Old Dominion University Norfolk, VA 23529
ABSTRACT
A neural network algorithm for speaker identification with large groups of speakers is described. This technique is derived from a technique in which an N-way speaker identification task is partitioned into N*(N-1)/2 two-way classification tasks. Each two-way classification task is performed using a small neural network which is a two-way, or pair-wise, network. The decisions of these two-way networks are then combined to make the N-way speaker identification decision (Rudasi and Zahorian, 1991 and 1992). Although very accurate, this method has the drawback of requiring a very large number of pair-wise networks. In the new approach two-way neural network classifiers, each of which is trained only to separate two speakers, are also used to separate other pairs of speakers. This method is able to greatly reduce the number of pair-wise classifiers required for making an N-way classification decision, especially when the number of speakers is very large. For 100 speakers extracted from TIMIT database, the number of pair-wise classifiers can be reduced by approximately a factor of 5, with only minor degradation in performance when 3 seconds or more of speech are used for identification. Using all 630 speakers from the TIMIT database, this method can be used to obtain over 99.7% accuracy. With the telephone version of the same database, an accuracy of 40.2% can be obtained. approach partitions an N-way speaker identification task with N*(N-1)/2 pair-wise classification tasks. Each of these pair-wise classification tasks is performed using a "small" neural network. Each of these pair-wise networks is trained to separate only two speakers. That is, each pair-wise network is trained using speech data from the two speakers for whom the network is intended to separate. The decisions of these pair-wise networks are then combined to make the N-way decision. For the N-speaker identification task there are N*(N-1)/2 pair-wise decisions. From these pair-wise decisions there are N-1 decisions which are relevant to a certain speaker. The relevant decisions for each speaker are then averaged and used as an estimate for the a posteriori probability of that speaker. The advantage of using this BPP technique relative to a single large neural network is that it significantly reduces the training time and requires less speech per speaker for training. The disadvantage of this technique is that it requires a large number of pair-wise classifiers. The purpose of this paper is to introduce a technique for reducing the number of pair-wise networks required by the BPP approach. This technique will also be referred to as reusable binary pair partitioning (RBPP). The remainder of this paper provides an explanation for this method and summarizes some experiments with the TIMIT and NTIMT databases used to evaluate this technique.

1.

INTRODUCTION
There are several well-established techniques for speaker recognition/identification (for example, see Gish and Schmidt for a tutorial article). Techniques include both parametric methods such as Gaussian Mixture Models (Reynolds and Rose, 1995) and nonparametric methods such as ones using vector quantization (Soong et al., 1985; Matsui and Furui, 1991) or ones which use neural networks (Bennani and Gallinari, 1991; Rudasi and Zahorian, 1991 and 1992). The neural network approach, which is used in this paper, although potentially very accurate, has the drawback that, when a large number of speakers (i.e., classes for a pattern recognizer) is considered, the training time required by the network becomes prohibitively long. Additionally, the required amount of training data becomes very large. For this reason, some investigators partition the speaker identification task into a number of small tasks. Each of these small tasks requires a small size network which can be trained in a shorter amount of time and with less training data (Bennani and Gallinari, 1991; Rudasi and Zahorian, 1991 and 1992). One of these partitioning techniques is called binary pair partitioning (BPP) (Rudasi and Zahorian, 1991 and 1992). This BPP
2. REUSABLE BINARY PAIRED PARTITIONING METHOD
The basis for expecting that pairwise networks can be used to separate many speaker pairs stems from the observation that speakers are likely to be clustered in similar groups. If a binary network is trained to separate two speakers, with one speaker from each of two widely separated groups, that network is also quite likely to be effective in separating the other speakers between those two groups. For example, if a certain network is trained to separate a specific female speaker from a specific male speaker, it is quite likely that that network will also separate many other female/male speaker pairs. To take advantage of this speaker clustering in a systematic way, we begin by arbitrarily selecting the first two speakers in our speaker population and then training a network to separate these two speakers, using only the available training data for these two speakers. This trained network is then evaluated as to how well it can separate all other possible pairs of speakers in our population, using the training data of these speakers. A trained network is considered sufficient to separate other pairs of speakers if its performance, on the training data of these pairs of speakers, exceeds a certain threshold. This trained network is then used to replace those pair-wise networks which would have been required by the BPP approach to separate those pairs of speakers. Thus the networks which would have been needed for separating those pairs of speakers are eliminated. We then train another pair-wise network which was not eliminated by any of the previously trained pair-wise networks. Then we use that newly trained network to eliminate other pair-wise networks as described above. This process of training a network and eliminating or replacing other networks is iterated until all pairwise networks are accounted for. In practice, as shown in the experimental section, this method can be used to greatly reduce the number of pair-wise networks that would have been required by the BPP approach. In our implementation of this method, each newly trained network is tested relative to all possible speaker pairs, including pairs for which there is an already trained network. If the newly trained network is able to better separate two speakers than the previously selected network, it replaces the previous network for that speaker pair. This process may also completely eliminate some of the initially trained networks. It also insures that the trained networks are used for best effectiveness and helps eliminate potential bias due to the ordering of the speakers.

In all experiments 20 cepstral coefficients (CC0 to CC20) were computed for each speech frame as follows. First, a second order high frequency pre-emphasis filter with a broad peak around 3 kHz was applied to the speech signal. The second step was to compute a 1024 point FFT from each 32 ms Kaiser-windowed (coefficient of 5.33) frame of speech data with the window advanced by 16 ms. The next step in processing was to compute the amplitude spectrum, logarithmically scale it, and then frequency warp it with bilinear warping function using a warping coefficient of .45. The next step was to compute the 20 cepstral coefficients as the cosine transform of the scaled magnitude spectrum over the frequency range 0 to 8000 Hz for TIMIT, and 300 Hz to 4000 Hz for NTIMIT.

Experiment I
This experiment was conducted to investigate tradeoffs between accuracy and number of networks needed as a function of a "threshold" parameter. Note that, in this application, since a neural network output of .5 implies no discrimination between the two speakers of a pair, whereas an output of 1.0 (or 0.0) implies perfect discrimination, it was very straightforward to define a threshold as some number between .5 and 1.0. In particular, the average neural network output level for all speech frames was compared to the threshold value to determine if a given network was suitable to discriminate two speakers. The experiment was conducted using all 102 speakers of dialect region 2 with the threshold value changed from 0.55 to 0.75 in steps of 0.05. For comparison the experiment was also conducted with the original BPP method, that is using all 5151 networks. Figure 1 shows the identification accuracy for these 102 speakers as a function of the amount of speech used for identification from each speaker and for different values of the threshold, and for the complete set of networks. The numbers of networks for the various thresholds were as follows: .55, 35; .60, 127; .65, 353, .70, 781, .75, 1686; BPP, 5151. The figure shows that when the threshold is higher than 0.70, there is no significant improvement in performance, and the performance is nearly equal to that of the BPP method.

3.

EXPERIMENTS
In order to evaluate this clustering method and compare it with the BPP approach, several experiments were conducted. The main goal of these experiments was to show that, for a large number of speakers, the RBPP method significantly reduces the number of pair-wise networks with very little degradation in identification accuracy. For all experiments each pair-wise network was a memoryless, feed-forward, multi-layer perceptron and was configured to have one hidden layer of 5 nodes and one output node. Backpropagation was used for training these networks with 200,000 network updates using an initial learning rate of 0.25. The learning rates was reduced by a factor of .96 every 5000 network updates. A momentum term of .6 was used. The TIMIT and NTIMIT speech databases were used for testing. These data bases each contain 10 sentences for each of 630 speakers sampled at a 16 kHz sampling rate. Five of these 10 sentences are phonetically-balanced sentences and are called SX sentences. Three of these 10 sentences are phonetically-diverse sentences and are called SI sentences. The other two sentences are dialect sentences and are called SA sentences. In all of our experiments seven sentences (5 SX sentences and 2 SI sentences) of each of the speakers were used for training and the other three sentences were used for evaluation. The NTIMIT data base contains the same data as the TIMIT data, except that all the speech materials were transmitted over phone lines.

This experiment was repeated using the same 102 speakers, but the NTIMIT data base. For this case the threshold values were varied from .55 to .70. The numbers of networks for the various thresholds were: .55, 156; .60, 691, .625, 1197; .65, 1849; .675, 2480; BPP, 5151. The most dramatic difference between this case and the data from TIMIT is that the overall accuracy is severely degraded, as expected. For this case, performance of the RBPP method is also approximately equal to that of the complete BPP method, for thresholds of .675 or more. However, many more networks are needed to reach this threshold than for the case of TIMIT.

Experiment II
120 100 80 % correct 60 40 20 0 0 2 4 6 8 10
10000
0.55 0.60 0.65 0.70 0.75 BPP
The purpose of this experiment was to experimentally determine the reduction in the number of required pair-wise networks increases as the number of speakers increases. We varied the number of speakers from 25 to 400 speakers. The number of networks required by the BPP approach are, of course, determined solely by the number of speakers. Figure 2 shows the number of networks required by each of the two approaches as a function of the number of speakers. A threshold of .70 was used for the RBPP method for the TIMIT data, and a threshold of .625 for the case of NTIMIT. As the figure shows, for the thresholds used, the number of networks required for the RBPP is approximately 1/10 of the number needed for the BPP approach for each number of speakers. Thus, the computational load of the recognizer is also reduced by about a factor of 10.

100000 length of test speech (sec) number of networks
90 80 70 60 % correct 50 40 30 20 10 0 0 5 length of test speech (sec)
Figure 1. Speaker identification accuracy as function of test speech length for various thresholds for replacing networks. The upper panel is for TIMIT, and the lower panel is for NTIMIT.

1000
100
0.55 0.60 0.625 0.65 0.675 BPP 10
10 10 100 number of speakers ti mi t ntimit bpp 1000
Figure 2. The number of networks required for the reusable binary paired partitioned classifier as a function of the number of speakers.

Experiment III
This experiment was conducted to evaluate the performance of the RBPP approach when a large number of speakers is considered. For this purpose, the RBPP system was trained for all 630 speakers for both the TIMIT and NTIMIT data bases. A threshold of .75 was used for the TIMIT data and .600 for the NTIMIT data base. A total of 43000 pair-wise networks were computed for TIMIT and 22,000 for NTIMIT as compared to 198,900 networks which would have been needed by the BPP approach. Figure 3 shows the performance achieved by the system for the 630 speakers as a function of the amount of speech used for evaluation. For each data base, two curves are drawn: one is the accuracy considering only if the correct choice, the other is the accuracy if the correct speaker is among the top 5 choices. For the 630 speakers of TIMIT, we obtained 96.4% accuracy when 3 seconds of speech (one sentence) was used. This accuracy improves to 100%, as the test length increases to 9 seconds and if a speaker is considered "correct" if among the top 5 choices of the classifier. For the case of NTIMIT, the corresponding accuracies are 25.6% and 65.7%. Compared to another study which reported speaker identification results for all 630 speakers of TIMIT (Reynolds et al., 1995), these results are very similar for the case of TIMIT, but lower for NTIMIT. However, it should also be noted, that are several other differences between our work and this previous work, including the fact that the two sentences with the same speech materials for all speakers (the SA sentences) were used as training sentences in this previous study, thus not making the tests totally text independent. In our work, the SA sentences were used for testing only, thus not biasing the training process.

The method provides a very convenient mechanism for trading off accuracy versus computational demands and storage requirements. If lower accuracy can be tolerated, a lower threshold for replacing networks can be used to reduce the number of networks needed. An apparent extension of the method would be to retain the reused networks, using all data from speakers in the two groups separated by the network. Presumably, this would further reduce the number of networks needed for a given level of performance. It could also allow better scaling properties to very large speaker populations.

4.

ACKNOWLEDGEMENT
3.

CONCLUSION
Portions of this work were supported by NSF grant IRI9217436.

The RBPP technique described in this paper was shown to be very effective in reducing the number of pair-wise networks required for an N-way speaker identification task compared to the BPP approach. The performance obtained with this technique is nearly as good as the BPP approach, provided the number of networks is reduced by no more than a factor of approximately 10.

5.

REFERENCES
120 100 80 % correct 60 40 20 0 0 2 4 6 8 10 length of test speech (sec) timit-1 timit-5 timit-1 timit-5
[1] Bennani, Y. and Gallinari, P. (1991), "On The Use Of TDNN-Extracted Features Information In Talker Identification," Proc. ICASSP-91, pp. 385-388. [2] Gish, H., and Schmidt, M. (1994), "Text-Independent Speaker Identification," IEEE Signal Processing Magazine, October, 1994, pp. 18-32. [3] Matsui, T. and Furui, S. (1991), "A Text-Independent Speaker Recognition Method Robust Against Utterance Variations," Proc. ICASSP-91, pp. 377-380. [4] Reynolds, D. A., Zissman, M. A., Quatieri, T. F., O'Leary, G. C., Carlson, B. A. (1995), "The Effects of Telephone Transmission on Speaker Recognition Performance," ICASSP-95, pp 329-332. [5] Reynolds, D. A., and Rose, D. A. (1995), "Robust TextIndependent Speaker Identification Using Gaussian Mixture Speaker Models," IEEE Trans. Speech, Audio Processing, 3, pp 72-83. [6] Rudasi, L. and Zahorian, S. A. (1991), "Text-independent Talker Identification with Neural Networks," Proc. ICASSP-91, pp. 389-392. [7] Rudasi, L. and Zahorian, S. A. (1992), "Text-Independent Speaker Identification using Binary-pair Partitioned Neural Networks," Proc. IJCNN-92, pp. IV: 679-684.

Figure 3. Performance achieved by the RBPP classifier for a 630 speaker identification task, as a function of the amount of speech used for evaluation.

--------------------------------------------------------------------------------

* Recognition korean speech coefficients noise
* Speaker effects identification sid channel
* Frequency vocalizations features african mel
* Speaker lda samples feature recognition
* Structure fine envelope amplitude temporal
* Phone conditions matching speaker approach
* Variability speech speaker sources channel
* Mllr principal component wpc variance
* Report colloquium speakers sign sheet
* Perception identification speech transitions steady
* Speech speaker channel usable yantorno
* Recognition speaker speech cepstral analysis
* Analysis lip shape intensity approach
* Variability speech sources factors context
* Berkeley network control neural line
* Speaker cation identi spectrum speech
* Speech research processing speaker yantorno
* Speech temple edu channel processing