Facial recognition in video forensics is the kind of technology that allows for identifying a human face by analyzing the facial features and comparing the virtual imaging against a database of photos to find a match.
Thanks to its widespread usage and implementation, it acts as an efficient deterrent of crime since would-be criminals are smart enough to realize their actions are being monitored. Should they decide to break the law, uncovering their true identity and prosecuting them would be easier than at any other point in the history of humankind.
Apart from its place in crime and digital forensics investigations, video facial recognition also shapes the future of our society by reducing queues and streamlining several essential processes in a variety of industries, most of which we’ve covered in this post.
If you stick around to the very end, we’ll also give you a glimpse into the other side of the argument by providing an overview of the most common concerns and criticisms surrounding the topic.
The early beginnings of facial recognition date far back to the 1960s. In that era, Woodrow Wilson Bledsoe started developing a system for classifying photos of faces. Although comparatively primitive compared to what technology can accomplish today, it was an important stepping stone to all that followed thereafter.
Fast forward to 2001, the technology was ready to be deployed at Super Bowl, which spared a huge controversy over essential human rights in the western parts of the world. This is also the year when facial recognition technology found its way into several law enforcement organizations.
Throughout the 2010s, several improvements had been made to it, and it even played a role in identifying and capturing Osama Bin Laden. Nowadays, you’ll find it almost anywhere, including in integrated operating system features like Windows Hello.
Use cases for video facial recognition
For it or against it, facial recognition is already a part of our civilization, at least in the modern parts of the world. As time goes on, it’s reasonable to expect that we will start seeing even more of it and the use cases are likely to grow in numbers and increase in diversity as well. To give you an overview of what it can do, the following examples should paint a good picture:
Biometric border checks
Inside the Schengen zone, no passport is necessary to travel from one country to the other. On one side of the coin, this means convenience – everyone eligible can enjoy hassle-free travel with virtually no restrictions. However, non-EU citizens still need to be tracked and have their passports stamped as they enter and exit the country.
To ensure no one is trying to cheat the system or overstay, the security officers need to make the necessary calculations manually, as well as go through the trouble of identifying each individual and authenticating their documents. Once facial recognition systems are fully implemented, this would, for example, stop someone from getting around the rules with a second passport as the biometric scan can identify this is in fact the same individual.
Airport boarding
Pre-pandemic, the airports used to be packed full of people looking to travel, and now the airline industry seems to be slowly picking up the pace again. This means there is some incentive to overhaul the existing security and passenger identification systems.
Since 2009, there has been some experimentation with using fingerprints as the biometric identification method of choice, but numerous issues kept holding it back from becoming the golden standard. For starters, there are 10 different fingers to scan, so which one should be standardized? And what happens in case an individual’s finger is injured or otherwise unscannable as is the case with prosthetics?
Quickly, it became clear that facial recognition was a much better alternative to fingerprint scans. Not only did it turn out to be faster to execute, but also far more accurate. Indeed, facial recognition search in video forensics makes it really simple to know who’s present at the airport at all times and allows the security personnel to identify and apprehend those who are trying to flee the country and escape the legal system.
And it benefits the passengers too – practical implementation has shown that facial recognition drastically reduces the boarding time due to streamlining the entire process. In concrete numbers, it tends to take no more than 2 seconds with a whopping 98% accuracy.
Event security
Facial recognition plays a major role in event security, particularly during sports events. If there’s a problematic individual present who tries to hide in the crowd, the technology can quickly pinpoint that person’s exact whereabouts.
At the same time, the very same technology can streamline the admission process, thus effectively killing two birds with one stone. In the future, we may even see the process being 100% digital without requiring you to present a physical ticket to enter the venue.
Finding a missing person
In China, there is a public surveillance system that ensures everyone’s safety. As you may know, facial recognition in video forensics helps digital forensic examiners identify exactly who’s present in the recording or a live broadcast thanks to cutting-edge digital forensics software solutions like VIP 2.0 by SalvationDATA. In case a CCTV unit picks up on a criminal activity taking place, identifying and apprehending the perpetrators is just a matter of time.
Since it’s capable of identifying the bad guys, the same can be said for any person, including those who have been kidnapped or become victims of human trafficking. The identity of the victims can be added to international databases, meaning that law enforcement agencies all around the world have an easier time spotting them and bringing them to safety as long as they’re spotted in a public space.
Tracking student and worker attendance
Certain universities make it mandatory to attend a certain amount of classes and they want to ensure the students are staying true to their commitment. In a similar fashion, the same technology can be used by employers to see who’s present at work and who’s missing.
Casino monitoring
There are more reasons than one why facial recognition is a lifesaver in the casino industry. The obvious reason is to catch cheaters and those who cause problems or are otherwise unwanted on the casino premises. Ideally, thanks to the advanced state of technology, such individuals can be identified and turned away at the door.
Another group of people the technology is suited for identifying is problem gamblers. On some occasions, those who realize they have a gambling problem would voluntarily exclude themselves from gambling to keep their problematic behavioral tendencies under control for the sake of their health. But sometimes, the inner impulses can get the better of them so they try to break their own commitment.
To keep the gambling environment safe for everyone, casinos would often share problem gamblers databases with each other. Once an individual has been identified as such in one casino, they may be barred entry when trying to enter other casinos. Facial recognition camera can even catch those who attempt to mask or disguise their appearance.
Service quality control
Facial recognition is more than just pinpointing someone’s true identity. In fact, it can also be used to examine someone’s facial expression to unveil that person’s emotional state. In Japan, McDonalds is studying facial expressions of their employees to see if they’re smiling enough.
The reason being is that smiling is associated with better overall customer satisfaction, which makes for a valuable indicator that helps the employers assess whether their employees are doing a good job.
Advantages of facial recognition
Facial recognition has the potential to change the world as we know it and the benefits are shining through in more ways than one:
Increasing security and deterring crime
Once would-be criminals realize they’re being monitored all the time with technology that knows their true identity, crime will decrease just on that basis alone. As the old saying goes, it’s the opportunity that makes a thief. If we take the opportunity away by implementing more public surveillance spots, the risk/reward would no longer incentivize criminal activity in a monitored space or environment, thus effectively deterring crime from happening in the first place.
Convenience
If facial recognition becomes integrated into all layers of our society, customer service booths could effectively become a thing of the past. Can you imagine paying without cash or a credit card on the spot by having your expenses connected to a central account from where you could pay for everything simply by scanning your face?
The concept otherwise known as biometric payments is already taking a foothold in China. Since it’s getting tons of customer praise, we can expect to see even more of it in the future.
Reducing processing times
No one likes standing in a queue, waiting to finally get their turn to pay up. Biometric payments has the power to drastically speed up the process. Since your purchase history would be directly linked to you as a person, this effectively reduces the risk of someone committing a crime by running away without paying for the goods.
3 core components of facial recognition in video forensics software
To better understand how facial recognition in DVR and video forensics software works, we’ll delve a bit deeper into the programmatic side of things and explain the complex technological processes in layman’s terms.
Detection
Before we can pinpoint someone’s identity, we need to detect whether there’s a human face in the picture to be identified. In day-to-day use, this is often seen when you’re trying to take a picture with your smartphone and the device automatically draws a box around someone’s face.
This is the very same technology that’s used in facial recognition in video forensics. From there, we can move to analysis, but not before we’ve established that there is a scannable face present or not.
Analysis
The next step is to generate a “unique fingerprint” by analyzing the person’s facial features. Here, the advanced digital forensics technology such as VIP 2.0 automatically measures the distance between the eyes, mouth, nose, and other distinguishing facial features as well as takes a look at what shape they represent.
Since it also measures their depth, someone’s face can be recognized from various different angles.
Recognition
Now that we’ve gathered all this data as part of the digital forensics investigation process, it’s time to make sense of it by matching it up with a photo database.
Often referred to as the “faceprint”, a digital forensics investigator would traditionally need to run it against a database and wait for a possible match. With modern face recognition technology, for example, Big Data Technology. However, this part of the process is completely automated.
These advanced facial recognition algorithms also have a place outside of crime investigation and security, with Facebook and its automatic identification of people in photos being just one example.
The ethical and privacy concerns of facial recognition
Despite all the benefits it can bring, not all are in favor of facial recognition software and its increasing role in our society. This is particularly true for the western parts of the world that tend to emphasize personal liberty and the right to privacy.
Below, we’ll give you a balanced overview of the criticism the use of technology has received over the years.
Constant surveillance and political oppression
Mass surveillance and personal liberty values tend to lie on the opposite end of the spectrum. Being able to stay on top of crime is one thing, but what happens if the technology is misused for things like political oppression and silencing of those who have a different view?
After all, the purpose of imprisonment and incarceration is to remove dangerous individuals from society as to protect the innocent, not to initiate a witch-hunt against political opponents.
Storing and processing personal data without permission
Facial recognition works in conjunction with machine learning. In other words, the algorithm is trained for better accuracy by going through massive data.
The problem is that, oftentimes, the photos used in the process were obtained from various places around the internet without asking the owner for permission to store them and use them for the purpose.
The potential for errors and mismatches
Although facial recognition is incredibly accurate (up to 98%), it’s not perfect as there have been cases where it got things wrong.
Consequently, someone could be prosecuted for a crime they did not commit. When we factor in how people can wear masks, grow facial hair or get a new haircut, the potential for making an error is even greater.
Privacy concerns
In the public space, filming and surveillance is generally allowed (with a few exceptions). However, certain people feel that being under constant surveillance is a violation of their human right to privacy.
This raises more questions than answers.
To what lengths are we as a society prepared to go secure people’s privacy and is treading on one’s privacy a worthy sacrifice to make in exchange for tighter security?
Balancing convenience and security against privacy concerns remains one of the primary ethical dilemmas surrounding facial recognition technology.
Conclusion
Despite the criticism, facial recognition in video forensics is a notable step forward in digital forensics as it gives us the ability to unmask a suspect’s true identity.
In turn, many otherwise unsolvable crimes are now solvable thanks to what the technology brings to the table. Plus, as we progress through time, it’s likely to influence our society in ways unimaginable in the present day.
