fB

It's been a long time since we've spoken. You can probably imagine what anyone who follows Iris wants to know. What can you tell us about the state of the LG--Iridian dispute?

LG

That it is still going on and that we remain confident in our position -- but I can't be much more specific than that.

fB

What can you tell us about the business side of things?

LG

Our business is robust --We're well ahead of last year--in all areas both private and public sector, in North America and all over the rest of the world. Big projects including India's Andhra Pradesh, and Canada's CATSA remain very active. And of course, we are actively pursuing opportunities relating to TWIC and RT, now that iris is cited as a complementary biometric in those programs.

In fact it is a really active time for iris, and especially for us.

1. LG 's very large job in Andhra Pradesh India is expanding. The number of enrollment locations has doubled to over 1200. And we just passed the 23 million persons enrolled mark, so with 80 million enrollments envisioned, we are about a third of the way to completion.

2. LG has been involved in the Iris On The Move platform prototypes built by Sarnoff.

3. The ICE Challenge is underway. And early results show there are some interesting possibilities.

4. Consolidation in the biometric space continues and the iris sector has not been overlooked. Securimetrics, a licensee with a niche market, was recently acquired by L-1 and surprisingly, for more than at least one financial concern thinks Iridian is worth.

5. Obviously, the reinsertion of iris in TWIC and in RT as an alternative biometric is welcomed. LG, with its integrated hardware-software solution has a big stake in the outcome, though I am sure other iris camera makers and their principal software supplier Iridian feel similarly. We think the fact card readers are built into LG IrisAccess 4000 platform is a plus for LG. The fact we work through HAZMAT gear should make us part of the solution for TWIC everywhere in some capacity, but especially in refineries, chemical plants, etc.

6. Iris will now feature in national ID in a G-8 country as the UK's recently announced national ID card program include iris. Dr. Daugman, the inventor of iris recognition, serves as a key advisor on technology to the UK Home Office. He is also participating in the ICE challenge with not 1 but 2 new algorithms in competition with his older work product.

fB

Speaking of John Daugman, the latest issue of BTT carried an announcement about Iridian "discarding" its Daugman crutches. Can you comment on this?

I can't imagine that Iridian is pleased by the BTT report that they have jettisoned the Daugman "crutches." Dr. Daugman's pioneering technology was critical to Iridian. John Daugman's fine work was and remains the very legs that allow Iridian to stand.

fB

But Iridian has announced a new algorithm that they say is dramatically improved -- in image capture for example, and also in improving false reject performance.

Good image capture quality is an area where LG excels. This facilitates interoperability and is critical in large-scale programs like Registered Traveler.

So I'm not surprised that they have taken steps to improve FTE (failure to enroll), particularly given what's at stake in the UK with the recently announced large scale program -- National ID.

FTE performance in the 2004 UK Iris test, -- as reported in The Times early that year-- was just dismal and ultimately required a major system retrofit by the integrator. Failing to improve their performance in this critical area would have been a fatal error. LG appreciates the importance of these features and you can count on LG being actively in the hunt in this next stage of the UK's iris program.

LG applauds any efforts to bring down FRR's (false rejection rate) to make the iris recognition experience as intuitively user-friendly as it can be. But our view is that FRR reduction isn't merely a software issue--particularly given the success we've had and the very low FRR rates we've achieved by focusing on optics and hardware design.

fB

What is OpenIris™ and what might that mean for the iris sector?

LG

LG success in the public sector and other iris recognition specialists success in defense-related applications with large databases have made clear that large users not only want -- but demand open architecture. LG has made "open architecture" a key part of its value and benefit proposition for a long time--whether focusing on front-end LG IrisAccess™ iris recognition systems or back-end LG IrisAccelerator™ high speed matching engine. LG's Canada airport win, and LG's selection for the Andhra Pradesh ration card entitlement program and criteria spelled out in other big projects in our pipeline only reinforce our belief that open architecture is the way. LG has yet to receive any detailed information on OpenIris™. It will be interesting for me personally to see how "open" the Iridian OpenIris™ system really is, particularly in view of Iridian's long-running commitment to a closed, proprietary architecture.

fB

What About Iris-On-The-Move®? Who is doing anything with it?

LG

The idea of how far away from the person being identified one can successfully employ iris recognition has been a key question since this technology was first commercialized with the introduction of LG IrisAccess™2200 in 1999. Sarnoff's prototype IOM models now identify moving subjects at 10 feet using LG iData™ software SDK. They also use iCAM 4000 series iris cameras from the third generation LG IrisAccess™ 4000 iris recognition system to enroll and to handle individual authentication in envisioned deployments. From my perspective, Iridian has long been singularly focused on iris software. So it isn't at all surprising that Iridian's press release says they want to be active in this area that represents a big competitive challenge, as it is where algorithm agnosticism may be gaining traction fastest.

fB

What exactly does the addition of all these new algorithms mean to the iris sector?

LG

Several things. First, there is a lot of interest in iris from many smart people. Second, while many new algorithms show promise, none have been field tested extensively, at least publicly. We wonder when an interoperable standard will emerge. For context, think about where fingerprints were only a few years ago. Third, there are options for people in the business, and for parties that want to get into the business. The expiration of the Flom-Safir patent in 2005 reduced significant barriers to entry. Fourth, barriers will continue to fall as patents on Iridian's Daugman algorithms expire in 2011.

Given the opportunities and the potential for new competitors to emerge, LG as market leader understands both the implications and opportunities new algorithms afford our company, the industry, and the world as a whole. We are aware of developments and follow them with interest.

We are eager to see what Iridian's reported departure from Daugman algorithms really means. Perhaps we will soon see a patent application for this new algorithm, a late submission to ICE, or even a reassurance that the new Iridian algorithm bridges nicely to all the existing Daugman algorithm-driven solutions in the field. And, of course, we are very interested to see what John Daugman comes up with next.

fB

There seems to be growing emphasis on the appropriateness of iris for national identity-related databases.

LG

This growing awareness stems from the fact that iris recognition is the only biometric technology initially designed to work in the 1 to n search environment. Nothing comes close to eliminating duplicate records with the speed iris does. LG has an exclusive feature that automatically checks databases for duplicate records, or records created under alias. This makes it impossible to enter even one keystroke on a new database record until this "check" [which takes about a second] is made. Rather than "scrub" fingerprint database records for dupes, iris would be an ideal choice, and LG's function can do the job quickly and accurately. One could argue that in the US TWIC and RT programs, iris should be the primary biometric -- at least when it comes to enrollment and ensuring there are no duplicates in the database. It would save lots of time and work.

Iridian says its new algorithm is good for national ID programs. But the Daugman algorithm built into our current solution has already proven itself in this application. LG Iris recognition hardware and software has already enrolled 20+ million persons in an Andhra Pradesh Ration Card Entitlement Management that will eventually have 80 million persons enrolled. 80 million persons! Andhra Pradesh would be the 15^th largest country in the world, if it were a separate country. So LG today offers a solution that already meets the needs of most, if not all nations. Particularly when run in concert with our IrisAccelerator high speed matching engine that, fully configured, delivers over 10 million matches per second.

fB

So with different algorithms on the horizon, is iris making any progress towards standardization and interoperability?

LG

LG is proud of the fact that LG IrisAccess™ 3000 and its successor IrisAccess™ 4000 systems won't allow anything less than a "GOOD" or "EXCELLENT" quality image to be enrolled. In developing a standard for robust authentication technology, LG believes that the interests of the technology and the public are best served by developing high quality common standards instead of determining some lowest common denominator level of performance.

We were surprised by a recent announcement suggesting some algorithms enabled iris recognition to be effectively employed with inferior quality images. The IT dictum "garbage in, garbage out" over-simplifies our reaction. But we are concerned that poorer images could impair countermeasure/anti-spoof functionality, as well as a need to see to see clearer, not cloudier criteria for standards, interoperability, image capture, and image storage -- things that parties like NIST, ICAO, and others continue to advocate.

While it is possible to create an iris template from a poor image, LG asks, "Why would anyone want this?" LG imagers are designed to work in tandem with software designed to support optics and hardware design. As long as security remains a key driver for selecting this technology, LG opposes quality compromise on imagers or images. Superior imagers are already available in the market, and the market has its own ways of sorting out which deliver and which don't.

fB

Last week's newswire also carried a story about Iridian licensing its IrisChip™ technology to India's 4G Informatics for development of a range of lower cost iris recognition cameras. What do you make of that?

Innovation efforts in iris technology always capture my attention. I have to admit that the selection of 4-G, for what is being portrayed as a huge step forward, was a surprise. I would have expected this kind of innovation not from an integrator, but from an optics/electronics manufacturer with a track record making iris recognition solutions work -- and possessing the wherewithal to invest in a venture like this. LG, Panasonic, OKI, and Securimetrics all have expertise building quality electronic products and have dealt with the optical, electronic, security, and human interface-related complexities inherent in building an iris recognition solution.

Maybe 4-G decided they wanted to build iris recognition products in India for that market as part of a strategy of producing less expensive cameras for sale in India. That is a worthwhile business objective, but a challenging one to be sure, given the design complexity and human factors issues involved. While price is a factor in purchase consideration, performance is what drives the sale. Time will tell, and those of us who have been around the sector know from experience how much time it takes to get it right.

fB

Anything else you'd like to add? Anything big on the horizon?

Peter, I appreciate the time you've taken. We are looking at having a larger story about what has transpired in India with Andhra Pradesh. It really is exciting how they are already building new e-government offerings on what we've been putting in place. It might not be the first place that comes to mind when you think of e-government, but it is exactly what they are doing there. Til then, it's always a pleasure.

back