Documentation :: FAQ

Thanks for checking out Face.com!

We've been getting a lot of questions, comments and general feedback through our forums, feedback mail, contact forms, and blogs. We thought it will be helpful to post some selected frequently asked questions, please feel free to continue and send us feedback and we will keep this page up to date!

1. What is face.com?

Face.com is a technology platform with best-in-class facial recognition software.  We use face recognition to scan your photos and search for faces that look like you and your friends in photos, so that we can help you tag, share, and find untagged photos of yourself and friends.

We offer two Facebook applications: 1.  Photo Finder for discovering photos of you and friends that are untagged and 2. Photo Tagger which makes photo sharing a breeze by grouping faces within an album, and allow bulk-tagging of your friends. Face.com also offers a robust, free, REST API to allow third-party developers to create their own original apps and services.

2. Why?

The idea of using facial recognition to help with photo sharing was spurred during geeky garage parties, where it seemed that while many photos are taken, only a few get discovered or shared because only a fraction got tagged. By the time we released Photo Tagger, over a billion photos were being uploaded on Facebook alone each month, and we wanted to help users share photos more effectively with their friends. Following our release of Photo Tagger we launched Face Alerts, a feature which alerts you whenever new photos of yourself are uploaded by your friends, even before it gets tagged. It was a huge hit by users who were interested of keeping track of which photos of themselves got uploaded.

As the face recognition technology got more common across the web for tagging people in photos, many developers approached Face.com in order to leverage our technologies to additional platforms such as mobile and other photo sharing sites. To answer this growing market need, and to drive developer creativity, we released our face recognition REST API, which let's you apply our technology to detect and recognize friends in your user's photos.

Our intent is to continue and leverage our face recognition platform to drive fun, friendly, useful services for consumers, either directly or though partners.

3. How does it work?

The Face.com platform uses our own proprietary technology to detect and recognize people's faces in photos. We have geared our technology towards high accuracy in everyday photos, and are able to scale it to billions of photos a month. The various Face.com services mash-up existing platforms such as Facebook in order to provide tagging, and as such requires your specific Login credentials in order to process photos within your friends social network. As a cloud technology, developers sign up for free and start incorporating the Face.com capabilities without installing any software, focusing only on building a great user-experience.

While we cannot reveal all our secrets behind our great technology, we regularly contribute work to academic research. If you're interested in learning more about the underlying algorithms head over to the Labeled Faces in the Wild, a great resource for most recent approaches to face recognition and where we publish our papers.

4. What about privacy?

Our services were built to provide fun useful user apps and services around photo sharing with your friends. We are hyper-sensitive about user privacy, and it's one of those things that has been engineered into every feature of what we do, and in what we don't do. The Face.com services rely on users to actively provide access to photos and tagging information so that Face.com will be able to produce results. Only information that was passed to Face.com will be used to deliver our services, and we'll only use your existing permissions and social relationships to do so.

For Facebook photos and users, Face.com uses Facebook’s platform APIs and the user's Facebook Login credentials, to mirror the Facebook privacy settings of you and your friends and gain access to information each time it is used. For example, if you chose to hide your facebook tags, our services will get blocked out when attempting to recognize you in photos. Same goes for specific privacy settings for photos and albums. As an extra measure, the Face.com platform cannot be used to access non-friend information, even if such information is accessible through the Facebook user experience.

For more on privacy and the different ways to control it, refer to our privacy policy page.

5. Can I find photos of a non-friend using their photo?

We actually get this question a lot. The quick answer is No, we only support recognition for friends and in people's photos you already have access to. However the real limit extends beyond any enforcement that we put in place - we only search in photos you provide us in the first place, either explicitly or through your Login credentials, and therefore can only look for faces in photos you already provided. If you create a false friend profile for someone, you will still only be able to search for them in photos that you have access to.

6. I'm a developer, how do I get started?

Head over to our getting started guide here, where you'll find pointers to more technology overview, guides, examples, and full API documentation. Our API Sandbox is a great place to quickly pick up our APIs and see it in action, check it out!

7. What can I do with your API?

While the most common uses we've seen are around photo tagging, we've seen creative applications that leverage face detection and recognition for personalization, mobile apps, search filtering and more. Check out our examples, and please let us know when you come up with something new - we'd love to blog about it!

8. What can't I do with your API?

You need to provide the Face.com API access to the photos for which you'd like the API to work on. In the case of social networks like Facebook and Twitter this means user-approvals on your service like a Login with Facebook dialog, and passing the session tokens as part of each API call. In other words you can only search for people and photos which you already have access to, your immediate friends list and their photos.

We are also limiting the number of photos you may process each hour. Finding yourself or friends on the open web will require you to scan the Internet for photos and is therefore not feasible. A more likely idea is to piggy-back existing text-search solutions like in the celebrityfindr.com example which uses the twitter search API.

Finally - you must agree to our terms of service, taking special note to user privacy.

9. What's the difference between Face Detection and Face Recognition?

Face detection allows you to locate faces in a photo and report back where those faces were found, without identifying features. Face Recognition attempts to guess the person in a photo based on additional information you provide during processing. For the full background head over to our Recognition How-To page.

10. Is your API free?

Yes, our API is free, requires attribution and is subject to Rate Limiting. If necessary, these limits are also expandable for free through our Whitelisting program. Go ahead and sign-up today!

11. Do you offer a standalone SDK that I can install?

Not at this time.

12. I have more questions - how do I contact you?

For technical questions and bug reports, our forums are probably the best place to start. For licensing and other queries try our contact form or email directly to feedback@face.com.

For the latest on Face.com, follow us on our blog, our developer news, and on twitter.