A S C R A P B O O K O F S O L U T I O N S F O R T H E P H O T O G R A P H E R
Enhancing the enjoyment of taking pictures with news that matters, features that entertain and images that delight. Published frequently.
6 May 2014
Image analysis and forensics is coming out of the lab with Izitru, a free app, Web site and API to test whether digital images have been edited. Izitru is pronounced "Is it true?"
Izitru is a product of Fourandsix Technologies Inc., co-founded by Hany Farid, Ph.D. and Kevin Connor, who worked on Photoshop at Adobe for 15 years. They explain:
At Fourandsix, we initially focused our efforts on developing tools for forensics experts. With Izitru, we're attacking the problem of photo authentication from a different angle. Here we allow anyone who captures a photo to certify the original before they make any modifications or begin sharing it. By starting with the original, we have a richer set of data to work with and we can apply more automated techniques to the analysis.
Of course, automated tests can never quite match the judgment of a trained expert with an expanded toolbox of forensic techniques, so we've included a "Challenge" procedure that can bring some of the more controversial images to our attention. Though we can't analyze every one personally, our goal is to apply our expertise where it can be most useful, while also identifying ways to continually improve the automated testing.
With affordable access to increasingly powerful image editing tools, distrust of photos is at an all-time high. At Izitru, our mission is to restore the persuasive power of photography by eliminating doubt.
Izitru does that by testing an uploaded image six different ways. Images that pass all six tests get the highest trust rating. It takes six seconds to generate an authenticated image link. The iPhone app include a camera app to authenticate images on the fly.
If you disagree with Izitru, you (and anyone else who see the shared image) can click the Challenge button. "Images that receive a high number of challenges will get a closer look from Izitru experts who will run more advanced tests. If appropriate, the trust rating will then be updated," the company said.
A TEST
We gave it a try. The site only accepts JPEGs for upload, so we went back to last year to dig up one of a statue. Uploading was slow and no batch option was provided.
The analysis step was pretty quick. It was just seconds before the result was displays on a page of its own.
Our score was "Medium Trust: No file modification detected." The result was explained below the image: "Though this image did not pass all of our forensic tests, we did not find compelling evidence that this is not an unmodified original file from a camera."
How accurate was that?
We clicked the link to explain how to improve our score and found a few things that might have lowered a score: the file had been resaved, it was uploaded from mobile Safari in iOS, the uploaded image was captured on a smartphone or tablet, the uploaded iPhone image was not taken with the Izitru app, you camera isn't yet completely recognized.
Each of these is explained in greater detail on the tips page.
We can't evaluate the last tip, although the report did recognize the camera correctly, but the first was certainly true. We had resaved the JPEG.
When we import JPEG images from the camera, we automatically run them through ExifTools to add copyright information. So while the image data wasn't manipulated, the Exif header was updated.
API
There's also an API that can be integrated into third-party Web sites to send images directly to the Izitru servers to confirm their legitimacy.
The Izitru Web site is online now and the app is expected to be released today on the Apple App Store. Both options are free. The API is fee-based.
More information is available in the news release below.
New Photo Sharing Service Izitru Establishes Photoshop-Free Zone
Web site, App and API Identify Unmodified Photos Using Cutting-Edge Forensics
SAN JOSE, Calif. -- Fourandsix Technologies, Inc. today has launched Izitru.com, a free image hosting site that applies advanced image forensics to prove that hosted photos have not been modified with Photoshop or other tools. Addressing a growing distrust of manipulated images, the new Web site and its companion iPhone app provide an easy way for consumers to share photos with credibility assured. In addition, a developer API offers the opportunity to integrate photo authentication into any Web site or to support enterprise workflows that demand reliable photo evidence.
Restoring Trust in Photography
Due to the wide availability of image editing software, it is becoming rare to see photos that haven't been modified in some way from the time they were first captured. Viewers are unsure of what to trust, whether they're looking at a selfie on Facebook, an item for sale on eBay or a dramatic storm cloud photo on Twitter. Notable photo hoaxes, sometimes even fooling legitimate news outlets, have heightened the sense of distrust to the point that even real photos are often questioned and picked apart online.
"While there's value in creative manipulation of photos, there remains a critical role for photos to play in documenting reality," said Fourandsix co-founder and CEO Kevin Connor. "Our goal with Izitru is to provide the anchor for that reality, so that the most trustworthy photos can easily stand out from the chatter of manipulated imagery."
Izitru integrates more than a decade's worth of image forensics research that can distinguish an original JPEG file captured with a digital camera from subsequent derivations of that file that may have been changed in some way. As soon as an image has been uploaded to Izitru, it is subjected to a series of six different forensic tests. Images that pass all six of these tests get the highest trust rating.
Share Trusted Images in Seconds
First-time visitors to Izitru.com can generate an authenticated image link in seconds. Just tap the Upload button and choose to upload any unmodified JPEG file from a digital camera. There's no need to create a user ID, though users who want to track and maintain multiple image links can sign in with an existing social media ID. The iPhone app further simplifies the process by including a built-in camera for capturing images.
Every image uploaded to Izitru is hosted on a dedicated page with a prominent trust rating. A short URL and convenient sharing buttons make it easy to share the result on social media or via email.
Expert Forensic Analysis When Needed
The automated forensic techniques used by Izitru are sufficient for most images, but there may occasionally be more challenging images that would benefit from a trained eye and more specialized analysis. For this reason, every image page on Izitru features a Challenge button that visitors can tap to indicate they disagree with the forensic analysis. Images that receive a high number of challenges will get a closer look from Izitru experts who will run more advanced tests. If appropriate, the trust rating will then be updated. This community feedback will also contribute to a continual refinement of Izitru's automated testing routines.
API for Integration with Web sites and Business Processes
Though the Izitru app and Web site are immediately useful on their own, the greatest potential of this new service lies with an API that will permit third-party Web sites and businesses to bypass Izitru.com and send images directly to the Izitru servers for analysis. Using this API, a social media site could automatically rate the trustworthiness of photos as they are shared or an insurance company could flag suspicious accident claim photos as they are submitted from a smartphone app.
Availability
The Izitru Web site is available immediately and the Izitru app will be made available today in the Apple App Store. Both are free. Direct use of the API for integrating Izitru with other Web sites or services is fee-based and interested parties should contact Izitru directly at contact@Izitru.com.
About Fourandsix Technologies, Inc.
Fourandsix Technologies, Inc. serves the mission of developing practical solutions that separate truth from deception in photos. It was co-founded in 2011 by two established imaging experts: Kevin Connor, a former leader of the Adobe Photoshop product line and Dartmouth College's Dr. Hany Farid, whose pioneering research in image forensics established the technical foundation of the company.