Facial recognition search: what are online photo search engines?
Image search services allow the user to reach a new level of Internet use. What previously could only be found through a verbal description can now be “explained” to a search engine visually. It’s not just about making online shopping easier or optimizing workflow. Human face recognition is now possible.
Today, face recognition technology (FRT) is popular in both the government and the commercial sector. For example, FRT is required for identification without a password (Face ID). In what areas is this technology used, why do we need online photo search engines, and what does privacy have to do with it? Read about it in this article.
Table of Contents
What Is Facial Recognition Search?
11 Facial Recognition Search Engines & Tools You Can Use Today
Google Image Search
PicTriev: Face Recognition
TinEye: Reverse Image Search
PimEyes: Face Search
Betaface
Yandex
Bing Image Search
Facebook
Pinterest Image Search
Social Catfish
Berify Reverse Image Search
Facial Expression Analysis
Facial Image Search: To Be or Not To Be?
Summary
Face Search FAQ
How do you do a face search?
Can you search for a face on Google?
How can I find someone using face recognition?
Does Google Image Search use facial recognition?
Is face search useful technology?
Can face search be used for privacy invasion?
What Is Facial Recognition Search?
Pattern recognition is an integral part of the human brain. The idea to equip machines with this function was based on the potential outcome. Face recognition tasks are now a significant area in the field of AI.
The FRT work algorithm consists of two stages: identification and verification. At the first stage, there is a definition of who the person is, and at the second, there is a confirmation of if this person is really who he or she claims to be.
Most engines work according to the following scheme:
- Face detection;
- Facial features detection;
- Face normalization;
- Feature extraction and descriptor computation;
- Verification.
The machine compares the received and processed information with a database to find a match with the “seen”.
A few years ago, this technology was used mostly for entertainment purposes, such as finding the name of the actor you can’t remember. The possibilities of facial recognition search do not end there. They also can be used to detect, for example, fake accounts in social networks. Or you can even protect yourself from fraudulent fundraising — the machine can help establish the real identity of a person whose photo was used by criminal elements to create counterfeit documents.
The commercial use of the technology behind facial recognition search is used in:
- Controlling access to objects or systems;
- Identification of violations;
- Determining the portrait of the buyer;
- Identification in the banking sector;
- Time management;
- Payment for services.
Municipalities can integrate facial recognition into smart city systems.
11 Facial Recognition Search Engines & Tools You Can Use Today
Today, the use of FRT is available to almost everyone. Next, we will look at the most stable services.
Google Image Search
Google’s free online image search service does not use face recognition in photo searches. However, it can help you find similar images. To start using Google Image Search, you need to find the search bar’s camera icon and drag your picture into this area. In addition, the picture can be downloaded from the computer by clicking on the icon. The third way is to paste the URL.
After receiving visual information, Google will create a mathematical model and find the most similar files.
PicTriev: Face Recognition
This service is also free, and it allows you to work with a database of celebrity faces. It selects people in the photo whose facial features have similarities and verifies them with existing images. Here, you can find your celebrity counterparts.
TinEye: Reverse Image Search
TinEye is one of the very first public search services. It uses reverse image search and can find modified photos if they have been cropped or color corrected. Search via image URL or download is available, as well as drag-n-drop. The search results can be sorted, for example, by the largest size or the best match.
PimEyes: Face Search
This is a relatively new service of European production, which immediately gained a good reputation. The site positions itself as a device for finding personal photos. It offers paid and free options. The free-of-charge usage allows you to establish the fact that your photos are on the Internet. A paid subscription unlocks all other possibilities, such as Deep Search, PDF generation, and results in online locations. Also, with paid plans, the site can send you notifications if it finds your photos in the index later.
Betaface
Betaface can be interesting for professionals and businesses. The product is not free, but it has a demo version. After uploading a photo from a device or a selfie, Betaface gives out matches in the photo and assumptions about additional information (age, level of attractiveness, presence of evening stubble, etc.). The site offers comparisons against celebrities or Wikipedia data and allows users to create a database.
Yandex
Yandex image search works on the same principle as Google. In 2020, a separate service for finding people by photo was moved to the general search area. The algorithm of the service is accurate enough to find a picture that is as similar as possible to a search query. There is also a downside: sometimes the number of results is so vast that it is difficult to find exactly what you need among them.
Bing Image Search
Bing was recently renamed to Microsoft Bing. The rebranding aims to highlight the achievements of the corporation, including the image search upgrade. This service can pick up a variety of images that picture people similar to you. The download mechanics are the same as for general search engines: Drag & Drop, download from device memory, or insert an URL. Bing is available to everyone at no cost.
Initially, the face recognition function was introduced to make it easier for users to tag friends in a photo. The social network itself determined which of the friends got into the frame, and the author of the picture could agree with this or re-mark people manually.
Facebook is currently using technology to keep its users safe. If the person’s image has been uploaded to the social network, they will know about it, even if they were not even marked.
The program processes each picture uploaded to the social network and determines if there are faces on it. A biometric map is then built, and the data is compared with information about other photos of faces.
Pinterest Image Search
With the help of Pinterest, you can find almost anything, including photos of faces of people who look like you — or even your own photos. Below each result will be a link to the file location.
To start using the service, you need to click on the magnifier icon, go to the search, and tap on the camera icon. After that, you can select the desired file from the gallery or take a selfie. The face recognition application, like the site, does not charge any fees.
Social Catfish
This service has a pleasant user interface and provides a reverse search service by photos, as well as by emails, names, phone numbers, and other parameters. The data is reconciled using information from social networks, dating sites, as well as scammer databases. Selfie upload possible.
The use of the service does not imply a demo version. However, you can pay for a 5-day trial period before signing up for a full subscription.
Berify Reverse Image Search
This tool does both reverse image and video searches.
The service imports your photos or videos to the site and then looks through its database and search engines. You can also get updates by email if new results come up.
Although the site’s services are paid, there is a free trial period for 3 days or 5 free searches after registration.
Facial Expression Analysis
The face expresses what, in many ways, makes us human — our emotions. AI technologies can also read (or at least try to) what emotion a person is expressing at the moment. This is possible because the work of facial muscles is associated with a specific emotion or a combination of them.
All emotions can be divided into six almost universal categories:
- Anger;
- Fear;
- Sadness;
- Happiness;
- Surprise;
- Disgust.
In order for the machine to draw conclusions about the displayed emotion, it takes the following steps:
- Face detection;
- Feature detection;
- Feature classification.
The data obtained is compared with the machine database. Searching for all indexed images online is not performed due to the specifics of the task.
Facial Image Search: To Be or Not To Be?
Every year, more and more people learn about face recognition technology. There are also new areas of application that go far beyond comparing your photos with the richest or famous. Computing power and databases are growing, and the methods of operation of the technology itself are developing and improving.
Nowadays, facial recognition is helping to identify scammers on social networks. In the future, private detectives and government officials will be able to do the same, if not already. It will also become more convenient for HRs to check the background of potential employees. Of course, entertainment goals aren’t going anywhere, and people will continue to search for leaked photos of their idols and other additional content online.
At the same time, any technology can be tried and turned against the creators, that is, people in general. So, for example, the creation of deepfakes caused a heated debate. It will be necessary to develop separate algorithms and, perhaps, even create new specialties to combat the harmful consequences of the existence of artificial photos or videos that look real.
The issue of privacy violation remains a hot topic. In the US, users of one social network were unhappy with the new function and sued the company. The same applies to sites for searching by photo: no one can say for what purpose the service’s capabilities will be used, even if the service presents itself as a harmless resource.
Of course, you can make your social network account private and not upload your photos anywhere else, but you can’t be 100% sure, right?
Summary
At this point, face search exists, and it is rapidly spreading. While some resources like Pinterest are more used for day-to-day activities, others can be useful for business and fact-checking.
Face Search FAQ
How do you do a face search?
In order to start using the face search, you need internet access and an image. Choose one of the services, for example, Yandex or Betaface, to complete the task, and go for it. Services are divided into paid and free.
Can you search for a face on Google?
Not really. Google can find images of people, but it does not perform face recognition; that is, it does not engage in identification. However, it can produce a large number of relevant results.
How can I find someone using face recognition?
For these purposes, there are many search engines, sites, and applications. Their accuracy may vary, but in general, they can be used to find specific information. However, there is currently no single database for face search for individuals.
Does Google Image Search use facial recognition?
Google has announced that its plans do not include the use of facial recognition technology in image search. Other companies have already set bad precedents because of FRT due to privacy concerns. However, a picture search result can be successful to a certain extent.
Is face search useful technology?
Face search and recognition technology are widely used in a variety of fields. Companies use it to control working hours, and cities are adopting it to build safer, more comfortable environments. In addition, entertainment purposes are also not in the last place in the purposes list.
Can face search be used for privacy invasion?
Any technology can, to one degree or another, violate privacy by falling into the wrong hands. Facial recognition technology actually collects biometric data without consent, so it can be seen as a violation of rights. Given that this data can be combined with other data, the answer is rather positive.