Facebook Knows Your Face

Introduction

Ever since April 2008, when Facebook surpassed MySpace in worldwide unique visitors, Facebook has continued to dominate the social media scene (Eldon). In fact, Facebook is the second most visited website in the world, second only to Google. Currently, Facebook receives over 23 billion views a day, blatantly overpowering other social networking sites such as Google+, Twitter, LinkedIn, Flickr, and MySpace (Top Ten Sites). One of the main factors that helped catapult Facebook into success is the ability for users to share albums of photos with their friends. This feature has been found to be extremely useful, allowing friends to share memories, news organizations to update followers with photos of traffic blocks, and for businesses to showcase new items. Due to Facebook's overwhelmingly large user base for uploading photos, they recently implemented a new feature - tag suggestions, hoping to make it easier for users to upload and tag people in photos and photo albums. Although this feature brings with it more intuitive and useful technology, it also brings a wave of worries on privacy violations.

Overview of Technology

According to an article in EWeek Magazine, over 100 million tags are added daily to photos posted on Facebook (Boulton). Tags, "are really nothing more than keywords used to describe a piece of data" (Chastain). In this document, tagging refers to adding information to photos, such as the people in the photo, or the location where the photo was taken. When tagging photos in Facebook, the process can become extremely time consuming, depending on the amount of photos that require tagging. For each photo the user uploads, they must go through, click on each person in the photo, and then type in their name.

Although this process has been used for years, Facebook recently realized the amount of effort required for users to tag people in large sets of photos. Photographers, who upload hundreds of photos at a time, were required to manually enter in the names of each person in each photo - a daunting task. Due to this, on December 15th 2010, Facebook announced that they would be implementing facial recognition software to make tagging groups of photos easier (Boulton). The facial recognition software would gather and organize groups of photos, isolating each person based on specific "facial landmarks". These facial features include, but are not limited to things such as the shape of your cheekbones, the length of your jaw line, the distance between your eyes, and the width of your nose (Bonsor).Then, the software would suggest who each person was, based on the uploader's friends list. This technique will allow users to upload photos en masse and only need to select that person's name once (Boulton). In order to incorporate this new technique of facial recognition, Facebook acquired Face.com - a company that specializes in pattern matching and has been working on a facial recognition API (Murphy).

Application

The user base for uploading photos on Facebook varies greatly. One of the more prominent groups has recently been targeted with the implementation of tag suggestions - photographers. Photographers use Facebook as a method to generate business. When a photographer uploads their most recent photo shoot to Facebook, it opens many opportunities to increase their customer base. When a photographer uploads a batch of photos, Facebook generates a related post, so that anyone who is "friends" with or "likes" that photographer, will see the new photos. Through the sharing of posts, photographers are likely to find new customers who are interested in their photographs. Using Facebook also provides the photographer with a distinct page where others can preview some of the photographer's work. This helps potential customers to view past photos through a standardized interface, providing them with more information on whether they would like to hire that photographer. Lastly, Facebook provides an excellent medium for photographers to communicate and exchange photos with their clients. Photographers can post images they've taken from a session and tag the clients in the photo. This not only makes the client happy, as they can see the photos without having to travel down to the photographer's office, but it also benefits the photographer. By tagging the client in the photos, the photos will now appear in the news feeds of anyone who is friends with the client, thus instantly referring the photographer (Digital Photography School).

Competition

Around March of 2011, news started to spread about a new smart phone application from Google - Google Goggles. Although the primary motivation of the application was to use the smartphone's camera to implement visual search and to read QR codes, it was soon realized that the application could also incorporate facial recognition. Due to this, Google announced that it would not pursue facial recognition in the smart phone application due to privacy concerns (Boulton). Google also stated that they would only begin to explore facial recognition in visual search when "a strong privacy model is in place", and that the service would be opt-in (Woollacott). Having an "opt-in" service would mean that all users would not be part of the facial recognition feature unless they specifically signed up.

Google's social networking site and current rival to Facebook, Google+, also implemented facial recognition for tagging photos. This feature, called "Find My Face", also scans over photos and suggests who should be tagged in the photo based off of previously uploaded photos. Google's "Find My Face" relies on the same underlying structure, but varies slightly on certain aspects. The main differences include that "Find My Face" is opt-in, unlike Facebook, whose users are all automatically signed up for facial recognition (Kincaid). Another key difference is that the automated tags generated in Google+ must be approved by the subject who was tagged, if the subject isn't connected to the user who created the tag (Goldman). Both differences were incorporated as a method to combat privacy concerns.

Privacy Concerns

Of course, with any new technology, there are always privacy concerns, and facial recognition is no exception. Both Facebook and Google have run into public unrest over the new facial recognition features and have worked to make them as safe as possible.

Due to the fact that Facebook is a closed network, requiring users to log-in in order to see content, privacy concerns have been less prominent. Being a closed social network means that not just anyone will see you as a tag suggestion - instead only those who are friends with you will. Facebook also announced that although the tag suggestion is turned on by default, all users can turn it off using their account's privacy settings (Oremus). Facebook isn't completely free of concern though, and many additions such as tag approval from those who were tagged, were added after public outcry.

Google appears to have received a more negative response to implementing facial recognition. Many believe that as Google Goggles, the smartphone application, continues to grow and expand, that facial recognition will allow others to simply take a photo of someone and be able to "Google Search" for all of their information (Woollacott). Since Google has the ability to crawl the web for searches, Google might be able to pull related images and information from Flickr or Picasa, gathering together a network of information and possibly revealing material that would have otherwise been unreachable such as that person's name.

Potential Future Uses

Currently, facial recognition tools are only being used to help aid social networking users to locate friends among photos, but there are many potential future uses for this technology.

At the 2011 Developer Forum in Beijing, Intel showcased many uses of facial recognition including targeted advertisements. The demonstration showed that various advertisements were shown depending on whether the camera detected a male or a female, and whether they seemed happy or sad ("Intel Labs Shows Future Applications for Facial Recognition"). This technology could be used in busy places such as malls or subway stations, where there are many people walking around in close proximity to advertisements.

There is also speculation on facial recognition software being able to pattern match and recognize specific items such as brands, or clothing. This could help further tailor advertisements not only in the scenario above, but also online. If Facebook used facial recognition software to match specific brands of clothing, they might be able to locate specific brands you wear in your photos, and tailor your advertisements to get deals on that specific brand of clothing (Murphy).

Conclusion

Whether we like it or not, facial recognition has begun to be a large part of the world around us. Like all things, there are both positives and negatives to this new tool. On one hand, facial recognition leads to a simpler process; it will be easier and quicker to tag albums of photographs, and possibly in the future, advertisements will be more tailored to suit your needs. On the downside, we are sacrificing a piece of ourselves and giving Facebook (and other technology giants) more information about ourselves. As with most things on the internet, it comes down to deciding if it's worth sacrificing some of our privacy for ease of use. This same concept has been seen countless times before - why did we sign up with Facebook and give them our personal information? In exchange to make it easier to communicate with friends. Why do people make accounts on YouTube and give them information on our age, email, and other preferences? In order to make it easier to find our favorite videos. Whether or not facial recognition is overall a "good thing" or a "bad thing", I'm not sure, but I know that we are the ones who control it. If we decide down the line that Facebook's facial recognition tools are too invasive, it's up to us to stop using it.