Smile You're on Camera! Controversies in Facial Biometrics

[Music]

[Music]

[Music] [Music]

[Music] so something you already are after you [Music] do I gain balance you know entity

[Music]

basically there's a right you know to travel these countries may be forced to stricter requirements okay I don't know the country nothing right so yeah yeah just an end to this country you know you know verifying the workers attend your work you know the lock into networks you know access security operating inside or one of my retirement [Music]

so all right so two types typically you know how how to sign something and they can be using to most to verify which is the one that we're very familiar with to authenticate and to identify so the first one is to verify that someone is with the same they are and another one is just to determine who the user is and this is a famous example in Las Vegas when they use it for web but anybody know what a will is a Las Vegas will tell you what a lot of a this whale is high roller a high roller right so [Music] so not only to know what happened so my ownership interests better than others

or some people and what was happening down you know so or he is right or can't strengthen their against because you know there are some challenges in Co usually have more pronounced for that but you know it's from seamless you know this technology so there's two types of identification systems positive and negative a positive identification system you want to verify that this biometric is in the database right this is kind of

[Music]

[Music]

[Music]

[Music]

you know because I observe this whole time limit you know using these technologies [Music] you know alright so looking at our neat characteristics on your face that will and bone structure knows ridges are this ordinate sizes and shapes and one thing about these technologies actually probably a very low resolution camera and facial stands it's already being done by the airports now now [Music]

so and one thing though about this recognition or I'm going to be this is the faces on the market is reliable way to identify someone from pictures from video from afar and Sergius link okay

and I have you that artificial geographies already being analyzed readily does your state government have your picture question it used to be world lapis smile doing your driver's license picture but now they don't seem to care about their by vectors or glasses or powerful [Music]

[Music] [Music]

[Music]

[Music]

but apparently they were able to create a system that would allow malware to target someone by their facial features [Music]

[Music] you know what all right so specifically and this is an automatic process of comparing two phases hey I'm gonna go to the process that you have to this process has to be that it's going to create some kind of a pretty ambitious and then quantify that numerically and then point you to compromise that America Merkley it's going to find it by the way this is not quite and they have from after when you log in you know your system password and it's really not quite so exact so really what its gonna do is going to try to identify those with the most similar features and it's not the same just like the regular [Music]

now now [Music] [Music]

[Music]

[Music]

[Music] [Applause] [Music] you are but I know some must have the body but our vendors so you know their technologies right you know the car we can track your phone right now they're able to track you there's no [Music]

[Music] yeah [Music]

[Music] [Music]

[Music]

[Music]

so they're probably insurance right now as we speak so there's four ways that law enforcement uses based identification to stop and identify this also personal identification so someone basically stopped they refused to identify themselves so essentially the Eugenia police officer takes their picture has some software or police car on some device and essentially checks it against mug shots and on a daily basis and determines you know who you are on that basis our rest on identifying this one here you're arrested okay so there's absolutely first one the resident and the mug shot that they take every other day [Music] and then investigating identify he here they basically have suspect you don't know why someone has actually done a

crime and what we do essentially is you know try to think that teenage of them and you know actually search to see what we have and then we also have a do something else this is what I was talking about that's becoming more more popular right essentially police will upload images a real-time video and essentially really concerning we're seeing a lot of growth in real-time we're also be seeing there was you know retail Walmart had a very failed example of this anybody know one retail would be interested in facial recognition you're right and I don't mention that in this presentation that we're working toward advertising there's a lot of models for that right now but there's a no reason

yeah cost prevention looking for shoplifters okay that wasn't really very successful so they do that uh how about airports you know essentially control

this kind of authentication what a comment on what that was Kim explain what happened

[Music]

Wow yeah exactly [Music]

[Music]

[Music]

[Music]

all right okay so also you deployed these are some samples

[Music]

[Music]

[Music] you know these technologies introns okay some of you but you know these jobs that contain yourself ease Cynthia send them out they can find anything a selfie well you know what might be a visual recognition from that perspective that's a consumer technology you know since you update a little bit and just want then you can have a drone going around on protest and you can have so you know they're probably a little lower low in the scale of this technology but of course I mentioned a border protection and this is kind of a big deal in our area but it's wrong [Music]

okay you know this was done all right [Music]

[Music]

and we turned it's trying it I'm not sure that's a call that I'm sure they don't want to try that but remember Skynet from the whole t2 movies my terminator is but this is actually a screen grab that my plan on YouTube on is just goes google it of how sophisticated their system is so this case here right they're looking at the photo in our right and then they are they to that the biometric detection on that picture here okay and then that basically I know I'm only Chinese able to be Chinese here but this should be some identifying information so so so what are the challenges are concerned to this technology well basically this technology creates

opportunities for tagging that other base by directors do not and me probably the most important I have an image secret about here we feel confident here that we're not being recorded people but I know that oh that when you're you changing your behavior changing [Music]

[Music]

[Music]

[Music]

[Music]

[Music]

[Music]

[Music]

[Music]

[Music]

[Music]

[Music]

[Music] [Music]

[Music]

[Music] but knows

[Music]

[Music]

you know for example without your permission and under the application of a precise content library it not only consider all details and all the new but including your name Facebook profile and [Music]

[Music] but these technologies grow this May it was discovered that over the last five years secret surveillance company founded by an Israeli intelligence officer had been quietly building a massive facial recognition database consistent basis acquired Facebook YouTube and countless other websites create these alarming privacy advocates and you know whether this is right you know a little bit to sociology so yeah you know how is she using it we don't know so my conclusions are this technology's you know under there that people's faces will eventually serve as their identification as opposed to try replace you

[Music]

[Music]

[Music] so

[Music]

[Music]

we're trying to force everybody to go they are go where the money is with it it is so but I also

[Music]

[Music]