However, there's been no comparable recommendation engine for works of art. If you like a particular Henri Matisse painting, there hasn't been a website for you to visit that will suggest other works by the artist or that will recommend other artists altogether.

But that's the aim of ArtFinder, a London-based startup that wants to help make it easier for people to find art that they love. Or rather, that's part of the aim.

A Last.fm, Shazam, Songkick, IMDb for Artwork

ArtFinder doesn't just offer a Last.fm-like recommendation engine, suggesting, for example, Paul Signac if you like Henri Matisse. It also offers an IMDb-like database of artwork and artists, displaying biographies and images. There are details where you can find a particular piece on display (that would be a Songkick-like feature, I suppose), and there's also a Shazam-like image recognition capability, so that you can snap a photo and get an ID and more information about a piece of art.

It's an ambitious undertaking, one that has to tackle the fact that a lot of galleries' and museums' art is not yet digitized and much of it restricted from being displayed online due to copyright and licensing restrictions. But that makes it all the more important as artwork is stuck in siloes - both offline and online - that make it inaccessible. The result isn't simply that many people think that art is "not for them," says co-founder Chris Thorpe. Art history books and curation efforts often assume a lot of knowledge, something that "puts people off" from discovering the art they love.

Helping Galleries, Museums, and Artists Build Apps

It isn't just a problem for art appreciation. It's a problem for museums and galleries, and it's certainly a problem for artists themselves. ArtFinder wants to be able to help people discover not just a love for Matisse, for example, but for contemporary artists who want to be able to sell their artwork.

ArtFinder is also helping museums and galleries build their own mobile apps, with a Wordpress-like platform meant to be easy for institutions to make their catalogs and digital assets available to visitors. This is where the image-recognition tool will be incredibly useful too - much to the chagrin, no doubt, of the docents who like to tell visitors to put their cameras away.

The first four of these apps are available now in the iTunes store: Cass Biomorphia, Cass Breaking The Mould 1 and Cass Breaking The Mould 2, and Watteau at the Wallace Collection.

Thorpe says that, until now, "we've never really had the right device" to bring artwork out of the galleries or out of the art imprints. But with the screen resolution offered on the iPad and on other tablets, you can now zoom into work to see the digital equivalent of the brushstrokes of a particular piece of painting.

Bringing this digital technology to the art world is an incredibly important endeavor when it comes to discovery, preservation, and commerce. Likening the process to the compilation album, Thorpe says that by building apps that showcase different "hits" from a particular collection, that the art world will hopefully be able to drive the equivalent of album sales - not just selling paintings and sculptures, but creating new art fans.

Want to try it out? Here's how. Live as of right now in the AppStore and the Android Marketplace is an updated version of junaio. In the app you can find the "junaio Glue" Channel either in the "New and Featured" category or in search results for "Glue" in the browser. Then, simply point your phone's camera at the superhero guy you see to the right (click the image for a larger version, it works better) and a 3D model with interactions and animations should appear. By pressing on the model, he will shoot his Metroid-like gun/arm, though no sounds are played when he does so. Also when I tried this on Android the model appeared upside down for some reason.

When I sat down with metaio co-founder and CTO Peter Meier earlier this month, he expressed to me the company's vision for mobile AR. While many mobile browsers have aimed to help people find things outdoors, Meier thinks the promising future of mobile AR is with indoor experiences.

"AR is super cool but not enough to change peoples habits - people aren't going to go outdoors more," says Meier. "Brand experiences are where the money is - sitting on a couch in a warm living room where no one is watching you hold your phone up."

With its Glue image tacking technology, the Munich-based company has a significant head start on other AR browsers. Layar - a browser with millions of users, over one thousand layers of content and a handful of partnerships with handset manufacturers - announced earlier this month that it hopes to include image tracking in its API and browser by the end of the year. By that time, metaio will have been providing mobile image tracking for as many as six to seven months.

Live image tracking on mobile phones is a likely catalyst for the proliferation of augmented reality advertising. Vendors like Total Immersion have seen excellent success and terrific revenues on desktop AR ad campaigns, and it would only make sense that these revenues will easily shift to mobile as the technology matures.

The battle seems to be which browser consumers will use to access the content, and Layar and junaio look like the current front-runners for browsers. Metaio will also include live tracking into its Unifeye Mobile SDK that lets third party app developers create their own apps independent of the browser - a competitor to Total Immersion's D'Fusion Mobile SDK. With its presence in both desktop and mobile based solutions, including both mobile browser and SDK offerings, metaio is in a good position to cover a lot of ground for potential AR marketing clients.

It's a bit of a strange work-flow compared to RedLaser's embeddable branded scanner on the iPhone, but it's a decent solution for Android developers. The app can also be launched from a mobile website by adding a special link in the HTML code of the site.

The SDK is free to use and supports several forms of 2D barcodes, including EZ codes, Datamatrix, QR codes and standard UPC barcodes. Right now the software development kit only supported on Android, but ScanLife says they are working on releasing an iPhone version soon.

It's great that more options are out there for developers to use emerging mobile technologies, but it would have been better to see ScanLife offer a more embeddable version of their scanner to foster more unified customer experiences. Having the ability to use the technology is one thing, but providing users with a smooth customer experience is a much higher value add for developers.

I uploaded a picture of my business card and Google Docs had trouble recognizing the largest text and clearest text on the card, but surprisingly did better with smaller text. A test of a PDF document turned up nearly perfect recognition results, but Google Docs strips nearly all of the formatting out, spewing out the text in a stream of letters and spaces. Other examples from Google Operating System produced decent results, but far from perfect or useful.

Additionally, when scanning a PDF, Google Docs does not save a copy of the PDF, so scanning to text and saving an original file requires two separate uploads. This feature is great for casual OCR users that want to quickly grab text from PDFs and some images or business cards. Those who rely on OCR heavily will likely be disappointed with the features and may have better results with commercial solutions.

The user placed objects will feature the ability to interact, connect, communicate and play with each other. Animated digital objects, which Layar is dubbing "Floaticons," can be placed in virtual space and fed with "health" to keep it around - increasing the likelihood of user interaction. Image recognition - a highly requested feature that could help bring desktop style AR to the mobile phone - is being developed in collaboration with Kooaba, a leading developer of the technology.

Matt Miesnieks, in charge of distribution at Layar, announced that LG, one of the world's largest makers of handsets, would be pre-loading Layar onto all of its upcoming Android devices. A few weeks ago, LG announced it would be stepping up its Android game with plans to release 20 new handsets worldwide by year's end. Layar has previously partnered with other manufacturers, like Samsung, and with this LG partnership the company will be on 1 in 3 new Android smartphones shipped this year.

Layar also announced some additional features that will soon be available on the browser, including background actions and flexible refresh rates for more dynamic layers and moving points of interest (POI). Additionally, developers will be able to fully customize their layer's AR experience by deciding what items to show and what to do when users click on objects. Finally, Layar will also now support animations and textures on 3D objects within layers.

Also of note, Layar has formed its first major brand partnership with the sandwich franchise Quiznos - a brand known for fun and edgy marketing. Users of Layar will be able to find restaurant locations through the AR browser, as well as view promotional 3D objects at various U.S. landmarks like the New York Stock Exchange bull, and the Hollywood sign.

Eventually, Quiznos hopes to allow for more functionality on Layar, possibly including ordering capabilities and location-based coupons. The partnership will benefit both companies, as Quiznos can extend its brand into AR and emerging social technologies, and Layar will likely benefit from further exposure via the Quiznos brand.

Another important partnership announced today is with National Geographic, which will be using Layar to create AR travel guides. The first one will launch in Washington D.C. and will let users see photos, 3D objects and other items to help them discover interesting facts about the city as they travel through it.

These announcements are encouraging to the greater augmented reality community as Layar makes progress not only within itself but with partnerships with brands and handset makers. According to Layar, it has 716,000 active users and will soon surpass 1,000 layers on its platform. Competition within the mobile AR browser space is heating up, which will drive innovation forward in the coming months and years.

For some AR experiences, image recognition and tracking are essential to a quality customer experience. The application needs to be able to analyze the data being taken in from the camera in order to properly overlay 3D objects into space, but previously this functionality wasn't available for iPhone developers. When Apple announced iOS 4.0, the inclusion of access to live camera data in the iPhone SDK provided AR developers with the ability to bring image tracking to the iPhone.

Other mobile operating systems, like Android, have had the capability to use image tracking for a while now, but the iPhone had been closed off from raw camera data. Last week, metaio introduced an image tracking application for Android, known as junaio Glue, and now with iOS 4.0, it will be able to bring this same functionality to the iPhone. The company also provides its own mobile AR SDK that allows developers to build apps leveraging metaio's technology, and Misslinger says metaio will include image tracking in the iPhone version of its SDK in the next few weeks.

New and Improved Cameras

But it's not just the software that will make image tracking on the iPhone easier and more available; the forthcoming iPhone 4 also includes a 5 megapixel camera capable of recording 720p video. The higher resolution of images captured by the phone makes the tracking of AR markers and image-based triggers much easier, but, as Misslinger points out, it comes with a catch.

Image tracking for AR requires that visual data be analyzed 30 times per second, and using a high-resolution image could slow down this process and make the tracking less accurate, or lag-ridden. To avoid this, Misslinger says the use of the full resolution will likely be on a case-to-case basis. For closer experiences, a lower resolution image should suffice, but for tracking markers that are larger or father away, the full power of the camera may be required.

The iPhone 4 also includes a second forward-facing camera, and everyone seems excited to see how Apple's vision for the future of video communication will play out. For augmented reality, however, the front facing camera opens up an entire new realm of possibilities on the mobile device. AR experiences traditionally developed for desktop webcams, like virtual mirrors that let users try on sunglasses and clothes, can now be experienced from a handheld device.

Gyroscopic Motion Sensing

One of the biggest surprises at the iPhone 4 announcement was the addition of a gyroscope to the device's arsenal of sensors. As Steve Jobs demonstrated, the gyroscope will allow the device to interpret its specific location as it relates to gravity - as the person turns in 3D space, the phone will recognize this motion based on sensed velocity. Most people immediately thought of augmented reality when this feature was introduced, and Misslinger says this is an obvious tool that AR developers will quickly adopt.

Certainly the gyroscope will help apps stabilize their results and track a user's movement, but there are additional uses for the gyro for AR. At the moment, if a user is using image tracking technology, quick movements of the device can cause the image to blur - interrupting the tracking of a marker or image. Misslinger says he is excited to use the gyroscope to help bridge the gap between when image tracking software loses and regains its capture on an image or marker. With the gyroscope working in tandem with image tracking technology, brief interruptions in tracking due to blurred images could be eliminated.

Faster Processor & High-Resolution Display

Call it what you will, but Apple's "Retina Display" on the iPhone 4 packs a serious high-res punch, and that bodes well for augmented reality. AR's success, in someways, relies on its ability to create a seamless merging of real and virtual worlds, and with a better display comes better graphics. Misslinger points out that the enhanced display will allow for the inclusion of better 3D models in mobile AR experiences - greatly enhancing the overall user experience.

Additionally, faster processor speeds on the device will allow these larger models to run much smoother than before. Apple's home-brewed A4 processor will allow AR apps to not only render 3D models faster and at a higher level of quality, but it will also help analyze the camera data at more close-to-real-time speeds.

iPhone 4 vs. The World

Apple is famous for admitting that it may not be first to include seemingly basic and simple functionality (like copy/paste and multitasking), but the company aims to do it as seamlessly and efficiently as possible. In the case of image tracking, Apple wasn't first to the game, but it is likely that with the ease at which developers can implement APIs from the SDK that the iPhone could soon establish itself as the leading platform for mobile AR. Add that to the phone's hardware additions and the large number of devices the company is likely to sell and you've got fertile soil from which AR can blossom.

Two such "divining" projects include CoPhIR (Content-based Photo Image Retrieval) Test-Collection and IBM's MUFIN (Multi-Feature Indexing Network). These projects tie into SAPIR's back end by extracting data from the Flickr archive and indexing features such as scalable color, color structure, color layout, shape edges and texture.

As shown in the video of Madrid's Plaza de España, SAPIR identifies matching media in the same way that humans derive intrinsic value from visual and sensory clues. Users can also choose to combine search terms with additional text to further drill down in search results. As is the case with regular search, if you already know the city where your image was taken, you're one step closer to finding your result. Additionally, SAPIR also has the ability to index sound and video files.

While the catalogue of media is still very limited, theoretically we may one day be able to search for almost anything using this technique. If Ashton Kutcher wears a pair of sunglasses we like, we can scan the image and search for the storefronts stocking them. If we're looking for the name of a town square, we can find it in the tags of similar images. And finally, if we're looking to self-diagnose we can compare photos of ourselves against jaundice or malaria patients.

The advantage of this tool is that we may one day have a chance to collect up the disparate bits in the digital ether and identify them as useful points of information. To test SAPIR in its early research stage, visit the homepage. You may also want to test out MUFIN to compare results.

Google is not the first company to add face recognition to online photo albums. Riya, when it first launched, made face recognition its central feature, but Riya's implementation, while interesting, made too many mistakes in the face recognition to be successful. Most users quickly abandoned the site after the early hype and Riya was forced to change its business model. To be successful, Google's face recognition will have to be a lot more precise than Riya's algorithms.

Picasa Update

The updates to the Picasa desktop software are interesting, but not revolutionary. Two features stand out, though. The movie maker mode will allow users to create videos of their slide shows and upload them to YouTube. The online synchronization feature will keep galleries in the desktop client in synch with the web albums. Sadly, though, this synchronization only seems to work from PC to web and not vice versa, making this potentially great feature just a bit less useful.

Competition

It's good to see that Google is showing some drive to innovate in the online photo album market. Flickr, as Stephen Shankland points out, has been slow to add new features, and Picasa itself has not exactly been the most exciting place to store pictures. If Google can make the face recognition work, it will have at least one great feature that will set it apart from its competition.

Windows Only

When Microsoft first publicly showed a demo of Photosynth in 2006, it almost looked too good to be true. Now, getting started with it couldn't be easier - as long as you have a Windows machine - there is no Mac version available yet. You first have to install both a browser plug-in and a desktop application (all done through just one installer). The installation was as standard as Windows installations get and finished in less than a minute. We tested the plugin in both Firefox 3 and Internet Explorer 7 without any problems.

One interesting aspect of the uploader is that you can choose a license for your creation. You can either choose a Creative Commons license, put the pictures in public domain, or mark it as 'All Rights Reserved.'

Desktop App

The desktop application does most of the heavy lifting for creating the 'synths' and seemed to make good use of all available cores. You only have to pick your photos, give your collection and name and click 'Synth.' After it has finished, it will create a score telling you how 'synthy' your photo collection was. Obviously, your photos need to have common areas for Photosynth to be able to stitch them together. While Photosynth does a great job making these connections, it can't work magic and our first attempts with relatively unconnected images were futile.

The more pictures you have, the longer the process of creating your synths takes, of course, and depending on your connection, the upload to Microsoft's servers can also take quite a while. In the end, though, your patience will be greatly rewarded.

Online Viewer

The online viewer is quite intuitive and allows you to zoom in and out, move around the picture and also go through the pictures in a 3D slide show mode. One nice feature is that you can also go full screen, which is really the best way to showcase your photos.

You can also embed a copy of your synth on any website and email a link to your friends.

The animation in Photosynth is astonishingly smooth and our screenshots really can't do it justice. If you haven't seen it yet, you should watch Microsoft's demo of Photosynth at TED2007, which will give you a good impression of what the final product looks like.

Different Way of Shooting

It really takes a different approach to shooting pictures to make the most out of Photosynth. If you often stitch together photos, you are probably already used to this, but Photosynth also gives you more freedom, as you can zoom in and out, or walk around an object and still have Photosynth recognize the common areas.

In our tests, Photosynth performed flawlessly, but we would recommend that you have a set of at least 10 to 20 photos to create an interesting 'synth' and the more photos you have, the more interesting it will be.

Caveats

A couple of caveats:

• All synths are public - there are no privacy controls!
• Photosynth only runs on Vista and XP so far.

Verdict

Even though we only had a short time to test Photosynth, it has already changed the way we think about taking pictures. Suddenly, you can do so much more with your photos. But besides the cool factor, we can also see a lot of other interesting applications for Photosynth. A realtor, for example, could use it to create a more immersive virtual tour of a house.

If you already have Photosynth installed already, you can see a 3D view of Venice below - otherwise, clicking on the image will take you directly to the installer.

From the labs of the makers of the image search engine TinEye comes an extraordinary product. The Multicolr Search Lab allows users to select up to 10 colors that they would like the images they are searching for to contain. To make a color more prominent in the image search results simply select that color several times from the provided color palette. As you pick and choose your colors the search results will automatically refresh with images that represent your new color selections.

So far, the Multicolr Search lab is available for Flickr and Alamy Stock Photography. However, not every Flickr image will be searched. According to the site, they extract the colors "from 3 million "interesting" Flickr images. Using our visual similarity technology you can navigate the collection by colour."

Unique Image Search Engines

In addition to the Multicolr Search Lab, I'd also recommend checking out the BYO Image Search Lab from Idée. The BYO or 'Bring Your Own' Image Search Lab allows you to upload an image or choose one that's available on the web and find images from Alamy with similarities such as color tones.

Both search engines are great and unique tools to make the best of. I didn't have any problems getting great matches to the colors and images that I provided. Another bonus is the use of the Flickr "Interesting" set to ensure the best quality photo results. These are two great services that we'll be sure to keep an eye on.

Nicholas Carr blogged about Gwap a couple of days after its launch, noting that "one thing the Internet enables, which wasn't possible before, at least not on anywhere near the same scale, is the transfer of human intelligence into machine intelligence." In Gwap, which stands for "Games With a Purpose," that transfer of intelligence is done by getting people to do the routine chores that computers don't know how to do - chores like tagging photos, describing songs, and outlining objects, as well as transferring a good bit of human common sense to the machine. The trick to getting people to do these things is to make the work fun. Hence the games.

The creator of these games is Luis von Ahn, winner of a 2006 MacArthur Foundation "genius grant" and a pioneer in the field of human computation. Ahn is most notable for helping to develop CAPTCHAs (Completely Automated Public Turing Test to Tell Computers and Humans Apart), those somewhat annoying but rather effective distorted letter puzzles used millions of times each day. Last year, he also introduced the "reCAPTCHA," where CAPTCHAs were used to gain access to a web site while also helping digitize old books.

Gwap homepage

The Games

Gwap currently features five games, one of which is an old classic called the ESP Game. In the ESP game, two players view the same image and try to guess words that the other player would use to describe it. Google licensed this technology and launched Google Image Labeler to help improve the quality of their image search results.

• Matchin, a game in which players judge which of two images is more appealing, is designed to eventually enable image searches to rank images based on which ones look the best.
• Tag a Tune, in which players describe songs so that computers can search for music other than by title - such as happy songs or love songs.
• Verbosity, a test of common sense knowledge that will amass facts for use by artificial intelligence programs.
• Squigl, a game in which players trace the outlines of objects in photographs to help teach computers to more readily recognize objects.

According to the Carnegie Mellon announcement, von Ahn plans to add a lot of games to the site, saying "we have three more that we'll be launching in the coming months." He hopes that by having all the games on the same site it will encourage players to try several different ones. Players also have a single sign-on and password, Top Player rankings, and online chats, said von Ahn.

The Human Processor

In his whitepaper entitled "Invisible Computing," von Ahn compared game design to to algorithm creation, saying:

"...it must be proven correct, its efficiency can be analyzed, a more efficient version can supersede a less efficient one, and so on. Instead of using a silicon processor, these "algorithms" run on a processor consisting of ordinary humans interacting with computers over the Internet."

In other words, we're the processor. The machine is us.

This concept isn't entirely new - Amazon's Mechanical Turk, for example, pays people to contribute their time to work on small, simple tasks called "Human Intelligence Tasks," or HITs. However, unlike HITs, which can sometimes be boring or tedious, the games on Gawp are actually fun - and they don't feel like work.

Some believe that human powered processing is the next big wave for computing. You could argue that Mahalo, the human-powered search engine is an example of this. (Though others call it a human-powered link farm.) Perhaps a better example is ChaCha, the mobile Q&A service that uses human guides to respond to questions called or texted in from your cell phone. We've also covered other human-powered services on RWW in the past, like the Galaxy Zoo and Stardust@Home project, among other (our coverage here). Many of these efforts have tried to incorporate an element of "fun" into what is actually work.

Whether Gwap will actually gain momentum and get a large number of people involved is yet to be seen, but it is definitely has potential to help teach computers the things they can't do for themselves....yet.

So a search technology that actually looked at a picture and could understand what was in it -- or at least match like images -- would have a large number of practical applications. Picollator's online demo is based on software of the same name from parent company Recogmission. The demo allows users to upload an image of a face to the site, and then attempts to match it to similar images it has gathered from around the web.

One of the main issues we've noticed with most of these facial recognition products is their inability to deal with my facial hear -- sometimes even thinking that my dark beard is really an indication of my skin tone. So I was excited when Picollator matched me to a picture of Russell Crowe. I don't really look like Crowe, but believe it or not, people actually used to tell me I did back in college.

Unfortunately for Picollator, subsequent tests with different images were less successful. It had trouble dealing with an older picture of mine (though to be fair it had less face and more body), and also didn't respond well to a black and white picture. And if their software really is accurate, I am dating a girl who basically looks like any celebrity that smiles.

What that means is basically that image recognition software still has a long way to go. Certainly, these technologies will improve over time, but for now, the flawed approach of text based image search from Google, Yahoo! and the other big search engines is probably the best way to go.