Webware

Google on Tuesday announced that its similar-images feature is now a standard part of the company's image search technology.

The feature was originally launched in late April, alongside the visual-news timeline, as a way for users to find images that share certain visual similarities with those in Google Images search results. This means that you could do a search for "ice cube" and very quickly fork out Google's results between images of the frozen chunks of water and the popular West Coast rapper, all without having to change your original search term.

Which 'ice cube' did you mean? Google's image search can now offer suggestions for images that may be similar.

(Credit: Screenshot by Josh Lowensohn/CNET)

Google continues to host a dedicated Similar Images search page that provides identical results to what's now found in the standard version of Google Images search. This is likely to remain, as has been the case for sites of other Google Labs graduates, such as Google Transit (now a part of Maps) and Google Suggest.

Google has also created a standalone Google Product Ideas page to grab user feedback for other features or changes to the image search service. This operates the same way as other product idea pages by letting users suggest new ideas to Google's engineers, as well as giving others the chance to vote ideas up or down.

Google image search is no longer restricted to iPhone and Android phones.

(Credit: Screenshot by Jessica Dolcourt/CNET)

Last March, Google's mobile team made it possible to search for images on your iPhone and Android phone. On Thursday, they rolled out image search for feature phones; that is, for pretty much any cell phone make and model with a Web browser.

Image search will work roughly the same way it does through a desktop browser. Starting from Google.com in the mobile browser, click "Images" in the top navigation, then fill in your search term. Google will return thumbnail images in the results page; between eight and fourteen, Google wrote in a blog post.

Click on one of these pictorial results to see a larger thumbnail image. Below, there's a link to view the photo in its original Web page and another one that takes you to the full-size image. You're also able to navigate "left" and "right" arrows to view other image results.

As with all Internet-dependent mobile apps, the slower your data or Wi-Fi connection is, the slower these images will be to load.

In an effort to keep people from incorrectly reusing or repurposing images found on its image search tool, Google has added new options that let users filter results by usage rights. Users can now filter photos by whether they're available for reuse, commercial reuse, reuse with modification, or commercial use with modification.

Google is including a variety of licensing methods including Creative Commons, GNU Free Documentation license, and items that are in the public domain. Its system for determining the rights on various shots is not foolproof though, and as such the company is recommending that those who are interested in republishing or reworking any of the images check with the content owner first (if possible).

Users can now choose one of four license filters for images on Google image search.

(Credit: CNET)

What's likely to be a long-term effect of supporting license filtering is that Google's image index becomes far larger than it is right now. Some content owners who have chosen to block its indexing to keep others from easily reusing their photos may think twice; with these new filters there's at least some semblance of care and control, even if Google is basing that off information from a photo's metadata.

It's also a signal that Google is paying more attention to the rights of user content, although the filtering is still something that's tucked away in the advanced settings of the search tool, and not something users have to check off before even beginning a new search.

Google has not yet rolled out license filtering to any of its other properties. Sites like Google Books and Video could be next. In the meantime, there are several search tools that let users quickly seek out images that can be reused and remixed including the Creative Commons search engine, Flickr, Blip.tv, ArtistServer, The Internet Archive, Wikihow, and Wikipedia.

Over the weekend Cooliris, one of my personal favorite Firefox add-ons released a really slick iPhone application that lets users search for images on Google, Flickr, Yahoo, SmugMug and DeviantArt just like they would on their computer browsers. The big difference is that you're simply able to swipe through the results with your finger, or tip your phone from side to side to navigate, which is wonderfully gratifying and natural.

Each search brings in the results in a three-image high wall that goes on nearly forever, and loads in as you continue to scroll to the right. Any photo can be zoomed into, and includes a link to the origin story, which can be opened up without kicking you back out to Safari.

Search Flickr, Google, Yahoo and others for photos on a giant, zoomy wall with Cooliris for iPhone.

(Credit: CNET Networks)

Besides the search tool, there's also an explore mode called discover that loads up the latest photos in one of five topics: election 08, news, sports, tech and business. These photos update throughout the day and are selected by where you are, meaning someone using the application in a different country is going to see a completely different list of items.

In future versions I'd like to see support for videos (like its desktop counterpart is capable of), however my most wanted feature is a way to save high quality copies of the images right to the phone. The built-in screenshot tool does a pretty good job until you want to start cropping. For a 1.0 product though, it's off to a great start.

Cooliris for iPhone is free. Embedded after the jump is a quick demo video of how it works.

Traditionally, Web search has relied on words or queries to scan massive indexes of pages for results. Searching for images can be a little trickier though. You're often relying on the competence of whoever uploaded the shot to provide the proper file name--and in a very small percentage of cases extra hints in the metadata to help the search tool get its hooks in.

In the real world, this isn't always the case, which is why the creators of a new image search tool called TinEye have approached image search the other way around--letting you search for sites using image files you've stored locally.

Sure this seems a little backward, but the idea is to find content related to whatever pictures you've got stored on your computer, or simply discover variations of that same shot around the Web.

Wondering where else an image has appeared? TinEye will let you know. In this case it's a shot from the Fail Blog (click to enlarge).

(Credit: CNET Networks)

The tool works best with popular or otherwise well-known images. Nearly everything else I tried didn't produce much. Some of the searches with the most results have been compiled in a "cool searches" section, but typically feature well-known art or photography. The company expects to have a better set of results as its index increases in size.

One of the more interesting uses for this technology is tracking down stolen intellectual property. We often find our posts put up on small blogs that cut out the bylines and take credit. If the blogs reused whatever screenshots we've added, a tool like TinEye would track them down even if they're using slightly altered text.

While we probably wouldn't go to such lengths to hound down a screenshot, there are photographers and other content creators who would. The company has already spun its technology off as a product called PixID, but it's geared more toward larger content rights holders than small-scale bloggers.

For users who want to spend a little less time (and bandwidth) using TinEye, there's a Firefox plug-in that will add a "search for image on TinEye" option when you right-click on any picture you come across on the Web. Users can also just grab the URL and enter it, which can be helpful if you don't feel like ferrying image files back and forth.

Several sites are already using the technology behind TinEye. Digg uses it to help search for duplicate story submissions on any post that contains an image. Adobe also uses it in PhotoShop Elements to look up related images that match the colors found in your photos.

Moving forward, the company intends to add video to its repertoire, letting you see where a video has been posted regardless of what service it's hosted on. This is something I'm far more interested in seeing than photos, as the viral spread of video clips across blogs and other pages is fairly rampant. When mixed with some sort of timeline, this service could yield some great metrics for video creators and a tracking system to follow when videos have been remixed and re-edited.

The service is currently in private beta, but we've got 100 invites for Webware readers who want to give it a spin. To get yours, fill out the Wufoo form after the break.

Google Image Search can now tell the difference between a picture that has a face in it and one that doesn't. But don't get too excited. it can't tell one face from another.

Step one: Do an image search

Technically, what Google can now do is similar to what some digital cameras do: Recognize faceness. Cameras do it to lock in focus; Google does it to separate pictures into "people" and "not people" buckets.

This feature is not in general release yet, but you can play with it. Here's how: Do an image search. For example, try a search on "CNET." You'll see a lot of pictures of products, buildings, and screenshots. Now go into the address bar and stick this on the end of the URL: &imgtype=face. Go to that URL and you'll see just CNET image search results of people.

You can also search for news results: Append &imgtype=news.

This isn't an official Google search feature yet, which is why you need to hack the URL to see it. Regarding the future for the capability, Google P.R. people only have this to say: "We're always working to improve the search user experience. In the near future, we will be launching a new search feature that will enable users to refine their searches for images pertaining to faces and news articles."

Step two: append the search URL with "&imgtype=face"

Result: Only pictures of people.

Historical note: A Web 2.0 startup, Riya, was pitching its facial recognition engine last year. It never worked well enough, and the company withdrew a lot from that product, launching instead a shopping search site, Like.com (preview) where you can select goods that look similar to the things you like.

Found on Google BlogScoped via Ars Technica.

The Zooomr photo-sharing site plans major changes, including the ability to let members sell their photos, but the upgrade process has been rocky.

(Credit: Zooomr)

Photo-sharing sites have added features such as tagging, commentary, ranking and printing. But adding the ability to sell photos injects a little profit motive in the business as well. It also puts the site in more direct competition with stock-photo sales sites such as Getty Images subsidiary iStockphoto.

Zooomr will keep 10 percent of revenue from photo sales, the company said on its blog, letting users keep 90 percent. For comparison, iStockphoto keeps 80 percent, unless users sign an exclusivity agreement under which the company keeps 60 percent. Getty also acquired journalism-oriented Scoopt this month, a site that splits revenue 50-50.

Zooomr went offline Tuesday for the overhaul to the Mark III version of the site. Users couldn't upload new images, though blog photos hosted on the site were still available. The upgrade was scheduled to be complete Thursday, but on Sunday, Zooomr rolled back to the earlier interface while debugging the new one.

"To keep everyone happy and continually uploading images, we've opted to put Zooomr Release Two back up temporarily for a few days while we get all of these Mark III bugs worked out of our system," site programmer Kristopher Tate said in a blog posting Sunday.

New search abilities also hampered the upgrade. The new version will let users search for images based on their colors, but even with five servers working flat-out, processing images for the search preparations took longer than expected, Tate said.

Zooomr today specializes in photo sharing augmented with support for multiple languages and geotagging, which maps the locations where photos were taken. The new Zooomr version also lifts file-size and quantity limits, the company said on its blog. And it will be endowed with a programming interface, allowing more sophisticated or automated interactions with the site.