start	end	text
1040	7420	Hi, my name is Camila. I'm a security researcher at DreamLab Technologies and also a 3D printing
7420	12540	enthusiast. And today we are going to talk about bypassing biometric systems
12540	16880	with and without the help of 3D printing technologies.
19040	26540	Humans have used features such as face, voice, and gait for thousands of years to recognize each
26540	32740	other. But only recently, humans have started using biometric-based systems to authenticate
32740	40820	individuals. Biometrics is the science of establishing or determining an identity based
40820	48380	on the physical or behavioral traits of an individual, such as the ones we can see here
48380	57660	in the slide. Fingerprints, DNA, signature, iris, face, voice, gait, vein pattern, ear shape,
57660	67270	keystroke dynamics, and more. Biometric systems are essentially pattern recognition systems
67270	74250	that read as input biometric data, extract the feature set from such data, and finally
74250	81290	compare it with a template set stored in the database. If the extracted feature set
81290	89270	from the given input is closed to a template set stored in the database, then the user is granted
89270	101760	access. But biometric systems are prone to different attacks. Direct attacks, presentation,
101760	109320	or spoofing attacks are performed at the sensor level. The sensor is full and not replaced nor
109320	116880	tampered. Indirect attacks are performed inside the biometric system by, for example, bypassing
116880	122920	or overriding the capture device, the signal processor, the comparator, or the decision
122920	130600	engine, manipulating the data in the biometric reference database, or exploiting possibly
130600	137620	weak points in the communication channels between the different components. But during this talk,
137620	148320	we are going to focus on this, on presentation attacks, this part of the system.
152080	159180	Now we are going to see presentation attacks in reality. These are real cases of criminals
159180	167060	using silicon masks to fool security cameras, airport security, and facial recognition systems.
167080	172700	The first one is a suspect in the robbery of a North Carolina bank.
174100	181920	This one robbed four banks and a CVS pharmacy with a silicon mask from an online site. I think the
181920	193460	of the mask name is The Player, from SPFXmask. It's a site that now is closed, but it used to sell
194820	203520	this type of realistic masks. The third one is a suspect accused of robbery and several
203520	211540	other crimes committed using a generic mask made by another online site. The name of this
211540	219820	of this mask model is The Neighbor. And the last one is a passenger who boarded a plane in Hong Kong
219820	232650	as an old man in a flat cap, but arrived in Canada as a young man. Regarding fingerprint biometric
232650	241110	systems, this first case is a Brazilian doctor who faced charges of fraud after being caught
241110	247730	on camera using silicon fingers to sign in for work for absent colleagues.
248270	256970	The second case is a gang involved in the illegal preparation, this is the image, the illegal
256970	263650	preparation and selling of cloned fingerprints to fool biometric attendance systems of several
263650	272090	educational institutions. The last case was in my country, Argentina. Six employees of the local
272090	279830	airline were fired after discovering that they falsified their entry to work with silicon fingers
279830	288330	taking turns to attend. But on weekends, when the payment is double, all six attended to work.
291740	300080	But how 3D printing could help bypass biometric systems? In this first case, researchers from
300080	310000	Forbes fooled Android facial recognition systems with a 3D printed head. In this second case,
310000	318660	the MSU Michigan State University team created a fake finger by 3D printing a mold.
321140	328120	And in this last case, a group of researchers from BCAB demonstrated that it was possible to
328120	335740	bypass the face recognition logging mechanism of the iPhone X using a cheap 3D printed mask
335740	346750	made from a stone powder. I love 3D printing, I have two 3D printers at home, why not make my own
346750	353910	experiments for bypassing biometric systems? But first, I wanted to try the traditional methods
353910	360050	for attacking biometric systems to better understand how 3D printing technology will help
360050	369880	make these attacks faster and better. But first, we need to know how is the fingerprint
369880	378870	recognition process. Most fingerprint scanners compare distinctive features of the fingerprint,
379120	386420	generally known as minutia. Typically, investigators concentrate on points where
386420	395440	rich lines end or where one rich splits into two called bifurcations. Collectively,
395440	405740	another distinctive feature, such as you can see here, delta, a short ridge, a spore,
405740	416880	the bifurcation, rich enclosure, a crossover or ridge, an island, all these features together
416880	428320	are called typica. The scanner uses complex algorithms to recognize and analyze these
428320	434980	distinctive features. The basic idea is to measure the relative positions of the features
434980	443960	in the same sort of way you might recognize a part of the sky by the relative position of the stars.
443960	452820	But to get a match, the scanner doesn't have to find the entire pattern of distinctive features,
452820	458280	both in the sample and in the print from the biometric database.
458280	464840	It simply has to find a sufficient number of features and patterns that the two prints have
464840	476050	in common. There are many fingerprint sensors on the market. These are optical sensors,
476050	484790	you can recognize them for the light, in general they use light. These are capacitive sensors,
484790	496010	and the last one is an ultrasonic sensor. Optical fingerprint sensors are the oldest method of
496010	502950	capturing and comparing fingerprints. This technique relies on capturing an optical image
502950	510790	and using algorithms to detect unique patterns on the surface by analyzing the lightest and
510790	523360	darkest areas of the image. Capacitive fingerprint scanners, instead of creating a traditional image
523360	530840	of a fingerprint, they use the body natural capacitance to read the fingerprints,
531480	533940	because the charge stored in the capacitor
535400	543140	will be changed slightly when a finger reach here is placed over the conductive plate,
543140	551080	while an air gap or a finger ballet will leave the charge at the capacitor relatively unchanged.
551480	558860	These changes can then be recorded and analyzed to look for distinctive and unique fingerprint
558860	568220	attributes. The latest fingerprint scanning technology is the ultrasonic sensor. To capture
568220	576060	the details of a fingerprint, the hardware consists of both an ultrasonic transmitter and a receiver.
576480	583400	An ultrasonic pulse is transmitted against the finger that is placed over the scanner. Some of
583400	589840	the ultrasonic pulse is absorbed and some of it is bounced back to the sensor, depending upon the
589840	596820	unique details of each fingerprint. The sensor then calculates the intensity of the returning
596820	604220	ultrasonic pulse at different points, resulting in a very detailed reproduction of the scanned
604220	613690	fingerprint. Now, for our test, the devices to be tested are four.
615050	619590	Two biometric attendance systems with optical sensors,
621530	628530	and two mobile phones, one with the capacitive sensor and the other one with an ultrasonic sensor.
634590	640710	This is the first photo of the materials bought for the experiments, and includes a lot of
640710	648150	materials. Alginate, a hot glue gun, gelatin powder, gummy bears, candle wax, transparent tape,
648150	655610	play-doh, instant glue, epoxy putty, UV resin, silicone fingertips, fingerprint ink, and more.
656250	662630	But during the test, I realized that I missed important materials, like, for example,
662630	669090	silicone, liquid latex, wood glue, and so much more, and the list grew, and grew a lot.
672620	679780	The first attack to test was the grease attack. For grease attacks, you need to have a clear
679780	688760	grease stain left on the surface of the fingerprint scanner, but this stain must have most of the
688760	694940	important features of the fingerprint left on the pad, so that the scanner can reliably read
694940	702120	the same line ends and curves that it detected on the previous user. The idea of the attack
702600	709120	is to gently press different materials, such as gummy bears, play-doh, silicone fingertips,
709120	716740	and latex gloves, against the fingerprint scanner, but with care, with careful, without
716740	729950	ruining the stain. Here are the results of the grease attacks. With gummy bears, play-doh,
729950	736210	latex gloves, and silicone fingers, the scanner detected a finger, but the fingerprint was not
736210	742490	clear enough to fool the sensor, so this attack was unsuccessful on all the tested devices.
743130	749470	But for me, this test was not a failure, because the gummy bears were really yummy,
749470	755750	and kept me fed during the rest of the experiments, so no failure for me.
758970	765020	The problem with grease attacks is that, in most cases, a regular grease stain
765790	772700	on the scanner surface is not enough to fool the sensor. We need to enhance it with other
772700	781880	substances to obtain better results, impersonating legitimate users. But this substance,
781880	790640	so that the user does not notice them, and also with ointment consistency, to better enhance
790640	799160	the fingerprint stain. This substance could be spread in the legitimate user fingerprint,
799160	808300	or in the fingerprint sensor. Using petrolatum ointment,
808300	813260	paraffin, or cocoa butter lip balm, we successfully fooled the sensors,
813260	819620	and were able to authenticate ourselves as the last user of the device, in optical
819620	834880	and incapacitive scanners. Now we are going to see a demo. In this case, we are using
834880	842180	cocoa butter lip balm. We are spreading the lip balm in the legitimate user fingerprint.
843600	849260	We can see the enhanced grease stain in the fingerprint scanner.
850860	858200	And then another user wearing a latex glove can be authenticated as the last legitimate user
858200	871240	of the device. Now for consensual attacks. The term consensual suggests that the user
871240	879600	we are taking the fingerprint from is aware of the process, and participates by pressing his or her
879600	889340	finger into some kind of a mold. For molds, we use these materials, alginate, epoxy putty, play-doh,
889340	895900	hot glue, and candle wax. And for casting, we use silicone, homemade ballistic gelatin,
895900	908130	liquid latex, synthetic resin, and wood glue. You can see here,
909010	919710	here, you can see here that the hot glue mold, of course, is the researcher's fingerprint are blurred,
919710	927710	but in this part, you can see that the hot glue mold is really, it's very detailed,
927710	941230	so it's a really good mold. For the molds, we obtained the best results with alginate
941230	949620	and hot glue. And for casting, we obtained the best results using liquid latex, wood glue, and silicone.
950450	958710	With the combination of a hot glue mold and liquid latex or wood glue casting, we were able to fool
958710	965530	all the sensors. The same with the combination of an alginate mold and liquid latex casting.
967930	975430	You can see here that the ballistic gelatin test didn't work. That's because ballistic
975430	981510	gelatin is not so easy to make at home. We tried several combinations of gelatin powder,
981510	986810	water, and glycerin, but the results were not enough to fool the scanners.
987610	995690	Also, note that the working fingerprints are very thin. And please be careful if you plan to make
995690	1004570	hot glue molds. Let the glue cool down a bit and test the temperature under the foil and dip your
1004570	1011870	finger in water because before, sorry, before pressing it against the hot glue, trust me, the
1011870	1020190	glue can reach over 200 degrees. I've been there and it burns and it burns a lot.
1024650	1033210	For unconsensual attacks, in these attacks, the user does not participate actively and latent
1033210	1041730	fingerprints are obtained in a non-cooperative way. Assuming we have identified the correct
1041730	1048570	latent fingerprint, we need to follow the following procedure.
1049690	1057430	This procedure here. So, we are going to need to enhance
1060050	1066590	the latent fingerprint with glue fumes or fingerprint powder, lift the latent fingerprint
1066590	1074030	with a digital camera or transparent tape, digitally enhance the fingerprints with software,
1074030	1080930	create a mold, and cast artificial fingers with silicone, liquid latex, or wood glue.
1081530	1090150	The first option to enhance the latent fingerprint is dusting with fingerprint powder and a brush.
1090150	1098810	The second option is encapsulating the latent print inside a container with instant glue.
1099030	1105690	Fumes from the glue will be attached to the wrist of the latent fingerprint, making it possible to
1105690	1119930	lift it. In this case, we obtain the best results lifting the latent fingerprints with a digital
1119930	1129030	camera, using a fingerprint enhancement software in Python to digitally enhance the fingerprint
1129030	1137810	image, offset printing a transparency, using the transparency as a mold, and casting it with
1137810	1146350	liquid latex. With this procedure, we were able to fool the optical sensors. The fingerprint ink
1146350	1154170	on a latex glove technique also worked on capacitive and ultrasonic sensors.
1155290	1166810	You can see here that the offset plate technique didn't work, but it did not work because
1167210	1174450	the offset plate was covered with some kind of rubber that is generally used in this type of
1174450	1181110	plate, and the rubber interfered with the creation of the mold. But without the rubber,
1181110	1195980	I think it's a technique that could work. For unconsensual attacks with 3D printing, we need
1196740	1204780	an UV Resin SLA 3D printer, software to digitally enhance the latent fingerprint,
1207520	1218160	a 3D CAD design tool, like for example Tinkercad, and a latent fingerprint in glass or a fingerprint
1218160	1228180	inked in paper. In this case, we can use FDM or filament 3D printers for these attacks because
1228180	1240470	we need the precision of an UV Resin printer. To obtain a working fingerprint through 3D printing,
1240470	1248130	we need to follow these steps. First, we need to lift the latent fingerprint with a digital camera
1248130	1255090	with macro functionality. Then, we need to use fingerprint enhancement software. In this case,
1255090	1263830	I use a software in Python, but you can use any kind of graphic software for this task.
1264310	1272150	Then, we need to import the enhanced image of the fingerprint into Tinkercad and configure
1272150	1280890	the dimensions and add the reach height to create the 3D model. One negative or hollow for casting
1280890	1290590	and one positive for direct tests. Then, we need to print the models on the 3D printer. In this case,
1290590	1299910	the Anycubic Photon 3D printer with UV Resin. And then, we need to use isopropyl alcohol and UV
1299910	1308890	post curing lamp or direct sunlight to complete the final curing process. At the end, we need to
1308890	1318150	cast the models with wood glue or liquid latex. It took us 10 retries to achieve the optimal printer
1318150	1327230	settings and reach height. But the most important step of this procedure is this one. It's the step
1327230	1336870	four. If the step four is okay, the fake fingerprints will work in the different
1336870	1347590	sensors and scanners. So, it's really important to configure the fingerprint length,
1347590	1358500	height, and the reach height in a correct way. The results.
1360220	1367920	The fingerprint obtained from the 3D model with liquid latex or wood glue casting worked on all
1367920	1376080	sensors. And the positive fingerprint print directly on UV Resin worked on the ultrasonic sensor
1376080	1379840	and in one of the optical sensors.
1381220	1387460	In the optical sensor, we had to spread the fingerprint with cocoa butter lip balm or
1387460	1397140	petrolatum for the sensor to recognize it as a finger. And here, you can see a summary
1397980	1405800	of all the process. Here is the fingerprint enhancement, the models in Tinkercad,
1405800	1413720	the printed models, and the casting with wood glue and liquid latex. And then, at the end, we are
1414280	1423000	using the fake fingerprints to authenticate in an optical scanner.
1424340	1436920	Now, we are going to see the demo of this attack. Here, we have the fake fingerprint
1438480	1449960	and we are using it to authenticate ourselves in a Samsung S10 phone with an ultrasonic sensor.
1449960	1458720	In this case, we use as a mold the 3D printing mold that we were talking before. And for casting,
1458720	1466900	we use liquid latex. In this case, we use liquid latex with a skin color, but it's not necessary.
1466900	1478650	You can use any color of latex and it will work. For biometric face recognition,
1479710	1487590	biometric face recognition is the process and ability to identify the face of an individual,
1488070	1495350	either to grant access to a system or to find out the details of a person by matching
1495350	1499470	the face with the data in the biometric database.
1499950	1509050	What a biometric face reader does is map and extract the distinct features, for example,
1509050	1517170	these points and these points here, features of a person's face that can be used for recognition
1517630	1525010	and stores the data in the biometric database along with the identity of the individual.
1525870	1533170	Our next step in this research is to perform presentation attacks in face recognition systems
1533170	1539030	by using 3D printed masks and heads to fool the different scanners.
1539370	1543050	Now we are working on that and it's really fun.
1545670	1551270	We published a paper of this research that you can download for more details.
1551270	1557470	Another interesting research was made by Paul Raskaniers from Talos. We met in Switzerland
1557970	1565130	when we were researching about using 3D printers to create fake fingerprints.
1565130	1571190	He used the same 3D printer for the test, but different software. Also, he tested different
1571190	1577310	devices, so if you are interested in this subject, it is worth to check it out too.
1577310	1580490	It's the first link of the reference materials.
1581270	1588710	Also, I have other reference materials in case some of the topics discussed in this talk
1588710	1593470	are of interest and you want to know them a little more.
1597120	1602400	Thank you. It was an honor to present this talk at DEFCON.
1602560	1608760	And also thanks to my co-workers and friends that helped me with this research,
1608760	1617760	especially Las Vivas Infosec. They are always there for me. So, stay safe, everyone.