Homomorphic Encryption Vs. Multiparty Computation

Bleeding edge encryption techniques allow one to monetize their personal health information without sacrificing privacy.

Encryption:

Encryption is the conversion of information into a code unreadable for unauthorized users by the use of an algorithm. Encryption is important when you don’t want anyone else to have access to your private data, such as brainwave data, selfie video data, or personal health data. There are many ways to compute (do math on) the encrypted data without knowing whose information it is about out of which, two are: Homomorphic Encryption and Multiparty Computation.

Homomorphic encryption is a kind of encryption in which the data is converted into a ciphertext which can be later analysed and worked on as it was still in its original form. The ciphertext is an encrypted version of the input data. It is operated on and then decrypted to obtain the desired output. This encryption allows us to perform complex mathematical operations on encrypted and secured data. It transforms one data set into another without harming the relationship between elements of both sets.

Multiparty computation is used to evaluate the inputs of two or more parties while keeping their inputs hidden from each other. This is done when different parties wish to jointly compute a function to their inputs in such a way that there certain security properties are preserved.

In simpler term, encryption allows us to hide data in a way that appears meaningless to anyone except those who have access to the secret decryption key. 

 

PROTECTION OF DATA THROUGH HOMOMORPHIC ENCRYPTION:

There have been many attempts to secure genomic privacy of biologically researched data using cryptographic methods. Particularly, it has been suggested that the privacy can be protected through homomorphic encryption.

The math on brainwave data recorded, of secret participants, using EEG while watching TV commercials, can be done through homomorphic encryption without decrypting the data.

The companies that get the brainwave data, never want to reveal the identity of their participants, that is why they send the samples in an encrypted form, to the lab, where the computations are done using homomorphic encryption, and the predictions (results) are sent back, to the company, in the encrypted form; where only they can decrypt it back using decryption keys. In this way, the identity of the person is never disclosed. The data is encrypted, also because companies and labs are bound by regulations and participant’s agreements to handle his data confidentially.

PROTECTION OF DATA THROUGH MULTIPARTY COMPUTATION:

Multiparty computation can be implemented using different protocols, such as Secret Sharing, in which the data from each party is divided and computed on separately. Then after combining again, it provides the desired statistical results. Security in multiparty computation means that the players’ inputs remain secured (except for the results that are computed) and the results computed are correctly. The security is supposed to be preserved In the face of any sort of adversary. Intuitively, no party learns about any other party’s inputs.

All in all, computation of encrypted data is an interesting topic that explains how cryptography faces the hardest problem of protecting data in use. This is just an overall review about what these two methods of computation have to offer us. The past few years have seen the most significant advances in making the use of these two technologies on more wider-scale.

 

For more information, see the project at OpenMined.org

 

 

References

BUILDING SAFE ARTIFICIAL INTELLIGENCE. (n.d.). Retrieved from open mined: https://www.openmined.org/

Computing Over Encrypted Data. (2017, may 29). Retrieved from Enigma: https://blog.enigma.co/computing-over-encrypted-data-d36621458447

homomorphic encryption. (n.d.). Retrieved from Search Security: https://searchsecurity.techtarget.com/definition/homomorphic-encryption

Homomorphic Encryption Market Size, Worldwide Analysis, Design Competition Strategies, Company Profile, Development Status, Opportunity Assessment and Industry Expansion Strategies 2027. (2018, june 7). Retrieved from 14 News: http://www.14news.com/story/38373250/homomorphic-encryption-market-size-worldwide-analysis-design-competition-strategies-company-profile-development-status-opportunity-assessment-and

Private predictive analysis on encrypted medical data. (2014, Aug). Retrieved from Science Direct: https://www.sciencedirect.com/science/article/pii/S1532046414000884

Secure multi-party computation made simple. (2006, Feb 1). Retrieved from Science Direct: https://www.sciencedirect.com/science/article/pii/S0166218X05002428

 

 

Alternative Modalities in BMI Biometrics

While most medical applications of the EEG BMI require clean data, recorded artifacts from facial movements and eye-blinks have been shown to be exceedingly accurate in classifying individuals.

Historically, when a subject undergoing an EEG recording session blinks or twitches their face, a sharp spike appears on the recording trace that renders that epoch unusable. While most medical applications of the EEG BMI require clean data, recorded artifacts from facial movements and eye-blinks have been shown to be exceedingly accurate in classifying individuals.

In addition to facial movements, eye-blinks are now being integrated with EEG systems to achieve a higher classification rate than using brainwaves alone. In [34], it was discovered that using a cheap, consumer-version EEG headset (the Neurosky Mindwave) achieved a 99.4% classification rate by discriminating eye-blinks in addition to EEG brainwaves. This represents the highest classification rate for a significant subject size (31 persons) to date using an EEG instrument.  Eyeblinks may have some disadvantages relative to brainwave patterns (such as forced-volition and lack of continuous verification), and are likely to be more prone to mimicry, but may be easier to train. Integrated with the continuous verification protocols offered by EEGs, eyeblinks may provide a high-level of security when requested in discretized intervals.

An exciting new modality analyzes differences in neural connectomes derived from functional magnetic resonance imaging (fMRI) to identify individuals. Unlike EEGs, fMRI relies on imaging deep neuronal structures, often in a clinical setting. The neural connectome is a diverse, complicated web of linkages among neurons inside the brain, which varies both with experience and genetics. In [35], connectomes constructed in the Human Connectome Project identified persons in a group of 126 persons with high accuracy (approximately 93%). It was noted that analyzing just the frontoparietal network was more distinctive across subjects compared to analyzing the whole brain (p < 10-9), and that these ‘neural fingerprints’ did not change when a person was at rest or performing a task. A person’s fMRI data from one session was used to create a connectivity matrix of their neural connections. This connectivity matrix was then compared to all other connectivity matrices from the remaining sessions, and a correlation coefficient for each session was calculated.

Disadvantages of using fMRI as a biometric is that the subject must remain stationary, it is contraindicated in subjects with implanted metallic devices, and it is significantly more expensive than other methods of biometric verification. However, conclusions drawn on fMRI data are likely to hold true on developing mobile technologies, such as DOT (Diffuse Optical Tomography) [36]. DOT has a greater temporal resolution than fMRI, allowing for continuous verification. However, it may be more dependent on detector placement than EEG [37].

 

Citations:

  1.   Facial movements, eeg authentication using artifacts http://link.springer.com/chapter/10.1007/978-3-319-07995-0_34#page-1
  2.   Multi-level approach based on eye-blinking http://www.sciencedirect.com/science/article/pii/S0167865515002433
  3.   http://www.nature.com/neuro/journal/v18/n11/full/nn.4135.html
  4.   DOI is good  http://www.nature.com/nphoton/journal/v8/n6/full/nphoton.2014.107.html
  5.   DOI vs fMRI – http://www.ncbi.nlm.nih.gov/pubmed/23578579
  6.   http://ieeexplore.ieee.org/xpls/icp.jsp?arnumber=1619442

Security Applications: Beyond Biometrics

A number of novel study designs have made use of the unique characteristics of BMIs to produce results that are not possible with other methods of verification.

The first example is Project Oblio. Project Oblio has an ambitious goal; to create a human-only area of the internet, beginning with a decentralized form of reCAPTCHA. If such a liveness detectable signal is also a biometric, then Project Oblio can create an anti-Sybil internet, with rate-limited cryptocurrency transactions at the user level, and ensure that everybody (even whales) have exactly one vote within Project Oblio’s government, prediction markets, and any other straw poll you might desire to post.

Another example is the bio-cyber machine gun (BCMG). This EEG-based password-validator works through a spin-off of the oddball paradigm, called the “spelling paradigm”. Letters that may be used in a password are grouped in regions, and a second set of letters are used to label these grouped regions. The region-letters (second set) are then flashed to a person wearing an EEG cap, in a random order. When a user is flashed the region-letter that corresponds to the region containing the desired letter in their password, their brain non-consciously emits a P300 brainwave, due to their underlying surprise or “peaked interest” correlated with the P300 inflection. (The P300 is commonly examined in commercial neuromarketing systems.) On the next go-round, a person is shown only the letters in their selected region, and can then choose the letter that comes next in their password. [24] Repeating this task, passwords can be strung together that have levels of entropy on a cryptographic level, as well as being tied to the biometric identity of a person’s brainwaves.

Another application involves EEGs in smart-home appliances for the disabled. In this set-up, visually-evoked potentials are able to both authenticate homeowners and reject nefarious individuals. An additional classifier based on imagined motor actions allow a disabled person to perform tasks, such as turn on and turn off lights, with moderately-high success (up to 85%). For those who may be quadriplegic, simple tasks such as turning a key in a locked door are impossible. Thus, a BMI set-up such as this provides not just security, but mobility, control, and independence, all with a single headcap. [25]

Brainwaves have also been used to generate a replicable PIN using a single-channel, commercial grade EEG [26]. Subjects underwent the oddball paradigm, viewing random presentations of digits 0-9. When the “password” digit was presented, a measurable brain wave called the P300 was produced, quite similar to the BCMG. Though the PIN was repeatedly classified with 100% accuracy initially, a latter publication by the same group indicated their classifier performance degraded each month of time following the training session (down to 78% for one subject after 3 months) [19].  This is one of the few studies looking at BMI biometric classifier degradation over time.

A protocol mentioned earlier utilizes recognition of EEG artifacts as a “covert warning” feature in the case of threat. The idea behind covert warning is that an authorized subject put at risk is capable of secretly broadcasting an alert that they are under attack, without alerting their attackers that they are calling for help. In this experiment, identified users wearing EEG caps clenched their teeth three times to produce sharp voltage spikes on the EEG trace, allowing for a signal to be detected 100% of the time. During this process, personal-identification rates dropped from 93% to 90%, a small drop-off considering the feature bonus. This is one of the only studies to expand on use cases for continuous verification. Notably, training data for this study was collected during various mental states (before and after caffeine intake, early and late in the day, etc.), which may have lead to its high classification rate. [14]

An increasingly  popular feature of security systems is the use of multi-factor authentication. Multi-factor authentication relies on a number of security measures, such as a fingerprint and a password, to authenticate an individual. Basic multi-factor authentication systems using BMIs have been reported in the literature [27]. However, BMIs are unique in that nearly any repeatable stimuli or task produces a distinguishable brain pattern. Thus, BMIs offer an endless number of “multi-task” authentication opportunities, all with a single headcap.

Quite recently, [28] proposed a multi-task learning system for BMI verification that interweaved information from finger-movement tasks to maximize learning. Subjects were asked to imagine moving either their left or right index finger. As was the case in [XX], the subject’s left side recordings were more distinguishable, but the greatest discrimination was obtained when using both the left and right side data together. These studies show that though some features are more reliable than others, integrating multi-factor authentication can produce even better security system than with one task, without the need for additional hardware.

Tasks that could be integrated into a multi-task authentication system include resting EEG state (including closed and opened eyes, see Table 1), imagined speech [10],  visually-evoked potentials for different objects (see [30], also Table 2), auditory-evoked potentials [29], solving a mathematical task [38], and imagining the rotation of an object or body part [13,38].  As implied, a user may choose which or however many of these stimuli to train their classifier on, adding a further security measure to this protocol.

Table 2: Classification accuracy of visually-evoked potentials using EEG.

eeg

 

 

 

  1.       http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4445586/
  2.       http://www.nature.com/nmeth/journal/v11/n6/full/nmeth.2936.html
  3.       http://www.ajol.info/index.php/amhsr/article/view/112158
  4.       https://www.cert.org/insider-threat/best-practices/
  5.       D&d – Need to find this citation still again
  6.       – Lost this citation too
  7.       http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=812278
  8.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=813403
  9.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=4107575
  10.   https://www.researchgate.net/profile/B_Kumar8/publication/224194495_Subject_identification_from_electroencephalogram_(EEG)_signals_during_imagined_speech/links/54a97f8a0cf2eecc56e6c45f.pdf
  11.  http://www.theregister.co.uk/2014/12/29/german_minister_fingered_as_hackers_steal_her_thumbprint_from_a_photo/
  12.  http://www.sciencedirect.com/science/article/pii/S0921889014002632
  13.   S. Marcel, J. R. Millan. Person authentication using brainwaves (EEG) and maximum a posteriori model adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 29(4):743–748, April 2007.
  14.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6199830&tag=1
  15.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7126357&tag=1
  16.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7314193
  17.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6994220
  18.  https://kar.kent.ac.uk/49535/1/PIN_ijcb.pdf http://www.inderscienceonline.com/doi/abs/10.1504/IJBM.2014.060960
  19.   http://www.researchgate.net/profile/Patrizio_Campisi/publication/282122165_EEG_Biometrics_for_User_Recognition_using_Visually_Evoked_Potentials/links/56039ec808ae08d4f171779c.pdf
  20.   http://www.sciencedirect.com/science/article/pii/S0925231215004725
  21.   http://link.springer.com/chapter/10.1007/978-3-319-19713-5_13#page-1
  22.  http://download.springer.com/static/pdf/440/chp%253A10.1007%252F978-3-642-27733-7_9145-2.pdf?originUrl=http%3A%2F%2Flink.springer.com%2Freferenceworkentry%2F10.1007%2F978-3-642-27733-7_9145-2&token2=exp=1447987756~acl=%2Fstatic%2Fpdf%2F440%2Fchp%25253A10.1007%25252F978-3-642-27733-7_9145-2.pdf%3ForiginUrl%3Dhttp%253A%252F%252Flink.springer.com%252Freferenceworkentry%252F10.1007%252F978-3-642-27733-7_9145-2*~hmac=be7c889f4ec177bd4096057e765bb2169f3614dfd4323d2e28787a2e3b92393c
  23.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
  24.  http://www.ncbi.nlm.nih.gov/pubmed/19163618   

http://www.igi-global.com.proxy.libraries.rutgers.edu/gateway/chapter/full-text-pdf/7471

 

  1.   http://www.metaljournal.com.ua/assets/Journal/MMI-6/060-Jianfeng-Hu.pdf
  2.   http://csee.essex.ac.uk/staff/palaniappan/1930378.pdf
  3.   http://ieeexplore.ieee.org.proxy.libraries.rutgers.edu/stamp/stamp.jsp?tp=&arnumber=5634487
  4.   – Need to find citation for this one as well  
  5.   Auditory potentials – https://etd.ohiolink.edu/!etd.send_file?accession=ucin1439300974&disposition=inline
  6.   http://ojs.bibsys.no/index.php/NIK/article/view/243
  7.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7280344
  8.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
  9.   Facial movements, eeg authentication using artifacts http://link.springer.com/chapter/10.1007/978-3-319-07995-0_34#page-1
  10.   Multi-level approach based on eye-blinking http://www.sciencedirect.com/science/article/pii/S0167865515002433
  11.   http://www.nature.com/neuro/journal/v18/n11/full/nn.4135.html
  12.   DOI is good  http://www.nature.com/nphoton/journal/v8/n6/full/nphoton.2014.107.html
  13.   DOI vs fMRI – http://www.ncbi.nlm.nih.gov/pubmed/23578579
  14.   http://ieeexplore.ieee.org/xpls/icp.jsp?arnumber=1619442

 

BMI Biometrics: What don’t we know?

Brain-machine interfaces are a biometric that's been untapped by leading governments and tech companies. Can we use anonymization techniques like homomorphic encryption to create computational barriers to fake-accounts?

Today’s EEG’s are commerical, having gelless electrode contacts for convenient recording outside the scalp. They are wirelessly compatible with mobile phones, cost far less than comparable biometric devices, and have established themselves as a novel consumer item. Given the unique advantages to BMIs already discussed (continuous verification, covert warnings, and inclusiveness), as well as their headcap design, EEGs could soon play a role in consumer virtual reality systems, as well as a corporate setting. Since many people could soon rely on the security of BMI biometrics, it is important that the major questions still unanswered in this underdeveloped field are brought to light.

While a large number of BMI studies involving EEGs have proven its ability to identify persons, there is a dearth in the field analyzing its potential for subverting security protocols. To date, only one study has evaluated security systems for storing EEG template data (see next section, or [16]), and only one has evaluated attacks on such a system (see [17]). This is probably because BMIs have yet to reach mainstream adoption, and there are no well-accepted protocols for BMIs in security systems. Once a standardized BMI security protocol is accepted, it will be easier to evaluate the BMI’s robustness in defense against spoofing.

Aside from the security issues, a number of basic, brain-based questions have  yet to be answered in the neuroscientific literature. The first has not been explored since the early, low-resolution studies. Can a brainwave identify differences between identical twins? This question is particularly relevant when looking at task-based biometric systems, which tend to have higher classification rates than resting-state studies (as were initially performed). Additionally, there are a multitude of effective mathematical models (Table 1) which could distinguish themselves on the basis of identifying identical twins.

Another concern on which research is sparse is that brainprints tend to change over a person’s lifetime. Initial research in this area suggests brainprints reach a mature, recognizable pattern shortly after puberty (age 19-20 years[6]), and become involuted with old age. In biometric studies, degradation of classifier performance due to this effect has been variable. Depending on the task used in the experiment, some classifiers have been found to degrade over a period of days [13] or weeks [18,19], while others have lasted up to 6 months [20]. In Marcel and Milan [13], there are indications that higher, long-lasting classification rates can be obtained when training data is collected over a period of days. A solution may be to have a short but infrequent training sessions over a week to establish a person’s identity, then update a person’s recording parameters each time they access sensitive data. Integrated with a password or “resting” EEG parameters, this may allow for effective updates to a person’s security parameters.

Two basic types of experiments exist in the BMI biometric literature: identification, and authentication. While identification experiments are more common, and may be a strong correlation for BMI’s authentication ability, they are far less practical in security settings. For BMIs, person-identification can be thought of as studies that pick out a person from a large group based on brainwave data recorded during a commontask. These experiments often involve recording resting-state brain features with no real active engagement by the user, and often have a lower classification rate. Alternatively, authentication protocols rely on a single task or series of tasks to identify a user. An authentication protocol should rely on a person using a series of “imaginations” known only to them (such as a password, mental image, or mental rotation of an object) to produce a distinct set of brainwaves. This person seems to ‘authorize’ a person based on this thought, regardless of how many members of the population make an attempt. Additionally, should a person be verbally instructed on how to make another’s “password thought”, the system should still reject this nefarious poser.

Though various types of “password thoughts” have proven effective throughout the literature (see next section), the next step to determine whether verifiable thoughts can be mimicked by nefarious individuals. One paradigm could involve a person viewing “live” brain recordings of both themselves and another, and attempting to alter their brain recordings to match that of the first person. The question of whether authenticating thoughts can be mimicked is particularly relevant in the case of identical twins. As a whole, the field of BMI biometrics must focus on developing studies that have “person-authentication” in mind, rather than person-identification.

Unfortunately, like biometrics as a whole, BMI authentication protocols may be at risk of a singular attack focused on obtaining or altering template parameters. To counter this, many biometric systems focus on key-binding architectures, which combine biometric templates with binary keys.  With properly chosen parameters, this protected setup approaches recognition rates close to those that are unprotected, and provides a quantifiable security-level of about 40 bits for this task.[14]

One long-held concern regarding the use of EEGs is the amount of time needed to train the classifier. In [31], it was noted that an increase in training time generally results in greater classifier performance. In this study, LVM classification had the highest classification rate (greater than 95%) after about 29 sessions, with each session consisting of 1-minute of recording (though interestingly, for a smaller number of training sessions, support vector mechanics was more effective). Although training sessions may be lengthy compared to other biometrics, [32] found that individual characteristics can be elucidated with an 88% classification rate based on only 0.2 second bins.

Lastly, a final concern of using BMIs as a biometric is privacy compliance. A brain recording may unveil personal health information of the recorded subject, such as a history of stroke or mental illness, epilepsy, or even alcoholism [21]. A severe trauma to the head or acute development of a neurodegenerative disease (such as a stroke) may lead to an unidentifiable brain pattern. Since BMI recordings are genetically-linked, they may one day be correlated with racial or other physical characteristics, permitting brainprints to identify an unknown individual.

Almost a decade ago, large-scale studies examining the nuances of a BMI-biometric protocol would have been too expensive to pursue. Today, there are numerous small-scale studies that approach BMI biometrics purely based on the potential of the underlying brainwave signal, and without regard to advances in encryption and decentralized machine learning. A number of replicable studies by those who recognize the potential for a one-person-one-vote internet and user-rate-limited, feeless transactions are needed before BMIs can be recommended for use in high-risk security settings.

Table 1: Person-identification in EEG biometrics and their classification. Taken from [22]

eeg1.png

http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985 [23]

.. others

Citations (posts continued in other pages)

 

  1.       http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4445586/
  2.       http://www.nature.com/nmeth/journal/v11/n6/full/nmeth.2936.html
  3.       http://www.ajol.info/index.php/amhsr/article/view/112158
  4.       https://www.cert.org/insider-threat/best-practices/
  5.       D&d – Need to find this citation still again
  6.       – Lost this citation too
  7.       http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=812278
  8.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=813403
  9.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=4107575
  10.   https://www.researchgate.net/profile/B_Kumar8/publication/224194495_Subject_identification_from_electroencephalogram_(EEG)_signals_during_imagined_speech/links/54a97f8a0cf2eecc56e6c45f.pdf
  11.  http://www.theregister.co.uk/2014/12/29/german_minister_fingered_as_hackers_steal_her_thumbprint_from_a_photo/
  12.  http://www.sciencedirect.com/science/article/pii/S0921889014002632
  13.   S. Marcel, J. R. Millan. Person authentication using brainwaves (EEG) and maximum a posteriori model adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 29(4):743–748, April 2007.
  14.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6199830&tag=1
  15.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7126357&tag=1
  16.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7314193
  17.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6994220
  18.  https://kar.kent.ac.uk/49535/1/PIN_ijcb.pdf http://www.inderscienceonline.com/doi/abs/10.1504/IJBM.2014.060960
  19.   http://www.researchgate.net/profile/Patrizio_Campisi/publication/282122165_EEG_Biometrics_for_User_Recognition_using_Visually_Evoked_Potentials/links/56039ec808ae08d4f171779c.pdf
  20.   http://www.sciencedirect.com/science/article/pii/S0925231215004725
  21.   http://link.springer.com/chapter/10.1007/978-3-319-19713-5_13#page-1
  22.  http://download.springer.com/static/pdf/440/chp%253A10.1007%252F978-3-642-27733-7_9145-2.pdf?originUrl=http%3A%2F%2Flink.springer.com%2Freferenceworkentry%2F10.1007%2F978-3-642-27733-7_9145-2&token2=exp=1447987756~acl=%2Fstatic%2Fpdf%2F440%2Fchp%25253A10.1007%25252F978-3-642-27733-7_9145-2.pdf%3ForiginUrl%3Dhttp%253A%252F%252Flink.springer.com%252Freferenceworkentry%252F10.1007%252F978-3-642-27733-7_9145-2*~hmac=be7c889f4ec177bd4096057e765bb2169f3614dfd4323d2e28787a2e3b92393c
  23.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
  24.  http://www.ncbi.nlm.nih.gov/pubmed/19163618   

http://www.igi-global.com.proxy.libraries.rutgers.edu/gateway/chapter/full-text-pdf/7471

 

  1.   http://www.metaljournal.com.ua/assets/Journal/MMI-6/060-Jianfeng-Hu.pdf
  2.   http://csee.essex.ac.uk/staff/palaniappan/1930378.pdf
  3.   http://ieeexplore.ieee.org.proxy.libraries.rutgers.edu/stamp/stamp.jsp?tp=&arnumber=5634487
  4.   – Need to find citation for this one as well  
  5.   Auditory potentials – https://etd.ohiolink.edu/!etd.send_file?accession=ucin1439300974&disposition=inline
  6.   http://ojs.bibsys.no/index.php/NIK/article/view/243
  7.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7280344
  8.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
  9.   Facial movements, eeg authentication using artifacts http://link.springer.com/chapter/10.1007/978-3-319-07995-0_34#page-1
  10.   Multi-level approach based on eye-blinking http://www.sciencedirect.com/science/article/pii/S0167865515002433
  11.   http://www.nature.com/neuro/journal/v18/n11/full/nn.4135.html
  12.   DOI is good  http://www.nature.com/nphoton/journal/v8/n6/full/nphoton.2014.107.html
  13.   DOI vs fMRI – http://www.ncbi.nlm.nih.gov/pubmed/23578579
  14.   http://ieeexplore.ieee.org/xpls/icp.jsp?arnumber=1619442

 

BMI Biometrics: What do we know?

Just like fingerprints, brainprints do have inheritable characteristics. However, unlike many other inputs to biometric analyses (fingerprints, irises, face-photos), brainprints have tremendous immunity to spoofing.

A brain-machine interface (BMI) is an instrument that records changes in electrical potential associated with thought, motor movement, or blood flow within a person’s skull [1,2,3]. As every person’s brain develops differently, these electrical potentials can be analyzed to distinctly identify persons with high accuracy. This is particularly relevant to the field of biometrics, in which biological indicators, such as fingerprints, faces, DNA, or brainwaves, can be used to discriminate between individuals. Today, biometrics are used for authentication when passwords may not be optimal, such as when someone may be “shoulder-surfing”, or if insider-threat level is high [4].  In BMI biometrics, the feature most often (but not always) studied is the “brainprint” — an underlying set of distinct features emitted from a person’s skull during a mental task. Given the remarkable applications of a one-person-one-vote layer to the internet, brain-based biometrics’ distinct advantages have high-potential applications.

It may come as a surprise that the earliest relevant work in BMI biometrics had the intention of finding genetic similarities, not identifiable differences. As early as 1936, researchers demonstrated that brainwave recordings were more similar among identical twins than among two random persons [5]. Three decades later, it was concluded that crude brainwave features have an autosomal-dominant method of inheritance [6]. It wasn’t until 1999 that the first person-identification study using brainwaves was published.  In (Poulos, 1999a), alpha-band rhythms of brainwave recordings identified individuals in a group with 72-80% accuracy [7]. It took only a few months for Poulos to apply a new method to the same dataset, improving his classification results rate to 95% through convex polygon intersections [8].

In 2007, the first large-scale BMI-biometric study was conducted by Palniappian et. al. By analyzing subjects’ brainwaves as they memorized images from a black-and-white picture set, a classification rate of 98.12% in 102 subjects was obtained [9]. In 2010, Brigham et. al applied support vector mechanics on brainwave recordings during subjects’ imagined speech, resulting in a 98.96% classification accuracy in 120 subjects, alerting the security community that brainprints were here to stay [10].

Just like fingerprints, brainprints do have inheritable characteristics. However,unlike many other inputs to biometric analyses (fingerprints, irises, faces), brainprints have tremendous immunity to spoofing. An attacker cannot synthetically generate BMI recordings to assume the identity of another individual as they can with a fingerprint [11], iris, gait [12], or facial recognition system. This is due to the tremendous complexity of the underlying biochemical processes of the brain and lack of public data available for training spoofers. When obtained in a human-detectable setting, brainprints may also be immune to stealing even from spoofing algorithms, unlike DNA. This may be performed in the short-term using vybuds as part of a “time-dependent input, biology-dependent output” communal hash, and in the long-term using OpenMined‘s mantra of “if good data were easy to fake, machine learning theorists would fake data instead of collecting it”.

In addition to BMI’s built-in “liveness” detector, BMI biometric systems are unique in that they require a user’s willingness to cooperate. Because brain patterns are heavily influenced by mood and stress-level [13,14], identification is not valid when forced by an outside party. This is a major benefit for user-safety. While this volition requirement is true for the vast majority of constructed paradigms for BMI biometrics to date, in some special cases (such as proving that a particular person committed a crime), a volition-insensitive BMI biometric may be desired. Initial research into developing protocols where identification is consistent regardless of a person’s stress-level has been explored[15].

Another advantage unique to BMI biometrics is their mobility. Recordings can be made unobtrusively, allowing for continuous verification throughout a secure task. This prevents one-time login attacks and person substitution. With continuous verification, a log-in cannot be transferred to others to allow them access to sensitive data. Given that digital security systems are most often thwarted by insiders [5], this advantage is particularly enticing for those working with corporate enterprise. Likewise, new research is taking this distinct feature of BMIs to the next level. Continuous verification allows for users to subtly initiate emergency broadcasts, without alerting a threatening individual who may be watching (see later section –  “Security Applications: Beyond Biometrics”). These “covert warnings” make use of distal artifacts (such as clenching one’s teeth) to secretly call for help without alerting one’s attackers that it has been requested [14], and are only possible through a continuous verification protocol.

A final dose of idiosyncrasy offered by BMIs is their inclusiveness. Those with severe injuries such as burned or missing fingers, aniridia (absence of the iris), or severe paralysis (such late-stage ALS or “locked-in” syndrome) are not excluded from using a BMI biometric. Due to their inclusiveness, BMIs are considered ‘universal’.

A number of different modalities have been used to record brainwaves, but to date, only two have been used in person-identification. The first, and most relevant, is the electroencephalography (EEG). The second, discussed later, is functional magnetic resonance imaging (fMRI).

 

 

  1.       http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4445586/
  2.       http://www.nature.com/nmeth/journal/v11/n6/full/nmeth.2936.html
  3.       http://www.ajol.info/index.php/amhsr/article/view/112158
  4.       https://www.cert.org/insider-threat/best-practices/
  5.       D&d – Need to find this citation still again
  6.       – Lost this citation too
  7.       http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=812278
  8.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=813403
  9.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=4107575
  10.   https://www.researchgate.net/profile/B_Kumar8/publication/224194495_Subject_identification_from_electroencephalogram_(EEG)_signals_during_imagined_speech/links/54a97f8a0cf2eecc56e6c45f.pdf
  11.  http://www.theregister.co.uk/2014/12/29/german_minister_fingered_as_hackers_steal_her_thumbprint_from_a_photo/
  12.  http://www.sciencedirect.com/science/article/pii/S0921889014002632
  13.   S. Marcel, J. R. Millan. Person authentication using brainwaves (EEG) and maximum a posteriori model adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 29(4):743–748, April 2007.
  14.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6199830&tag=1
  15.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7126357&tag=1
  16.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7314193
  17.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6994220
  18.  https://kar.kent.ac.uk/49535/1/PIN_ijcb.pdf http://www.inderscienceonline.com/doi/abs/10.1504/IJBM.2014.060960
  19.   http://www.researchgate.net/profile/Patrizio_Campisi/publication/282122165_EEG_Biometrics_for_User_Recognition_using_Visually_Evoked_Potentials/links/56039ec808ae08d4f171779c.pdf
  20.   http://www.sciencedirect.com/science/article/pii/S0925231215004725
  21.   http://link.springer.com/chapter/10.1007/978-3-319-19713-5_13#page-1
  22.  http://download.springer.com/static/pdf/440/chp%253A10.1007%252F978-3-642-27733-7_9145-2.pdf?originUrl=http%3A%2F%2Flink.springer.com%2Freferenceworkentry%2F10.1007%2F978-3-642-27733-7_9145-2&token2=exp=1447987756~acl=%2Fstatic%2Fpdf%2F440%2Fchp%25253A10.1007%25252F978-3-642-27733-7_9145-2.pdf%3ForiginUrl%3Dhttp%253A%252F%252Flink.springer.com%252Freferenceworkentry%252F10.1007%252F978-3-642-27733-7_9145-2*~hmac=be7c889f4ec177bd4096057e765bb2169f3614dfd4323d2e28787a2e3b92393c
  23.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
  24.  http://www.ncbi.nlm.nih.gov/pubmed/19163618   

http://www.igi-global.com.proxy.libraries.rutgers.edu/gateway/chapter/full-text-pdf/7471

 

  1.   http://www.metaljournal.com.ua/assets/Journal/MMI-6/060-Jianfeng-Hu.pdf
  2.   http://csee.essex.ac.uk/staff/palaniappan/1930378.pdf
  3.   http://ieeexplore.ieee.org.proxy.libraries.rutgers.edu/stamp/stamp.jsp?tp=&arnumber=5634487
  4.   – Need to find citation for this one as well  
  5.   Auditory potentials – https://etd.ohiolink.edu/!etd.send_file?accession=ucin1439300974&disposition=inline
  6.   http://ojs.bibsys.no/index.php/NIK/article/view/243
  7.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7280344
  8.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
  9.   Facial movements, eeg authentication using artifacts http://link.springer.com/chapter/10.1007/978-3-319-07995-0_34#page-1
  10.   Multi-level approach based on eye-blinking http://www.sciencedirect.com/science/article/pii/S0167865515002433
  11.   http://www.nature.com/neuro/journal/v18/n11/full/nn.4135.html
  12.   DOI is good  http://www.nature.com/nphoton/journal/v8/n6/full/nphoton.2014.107.html
  13.   DOI vs fMRI – http://www.ncbi.nlm.nih.gov/pubmed/23578579
  14.   http://ieeexplore.ieee.org/xpls/icp.jsp?arnumber=1619442

 

Brain-Machine Interfaces In Biometric Verification

Brain-machine interfaces (BMIs) show continued promise for applications in neuroprosthetics, neurogaming, and brain-to-brain communication. One area of BMIs often overlooked is in biometric verification. Biometric verification relies on identifying persons based on their biological traits such as fingerprints, irises, faces, etc. Using either EEG (electroencephalography) or fMRI (functional magnetic resonance imaging), BMIs have shown to be effective in distinguishing the identity of subjects with a near-perfect classification rate.

To date, there have been a number of BMI biometric studies that make use of identifiable brain patterns to classify individuals. BMI biometric protocols offer idiosyncratic advantages to comparable biometrics, including a volitional requirement, continuous verification, covert warnings, universality, and multi-task authentication. However, a number of basic security questions remain unanswered in this young field. Can authorizing thoughts be mimicked? Are brain patterns of identical twins distinguishable? Is there a way to ensure brainwaves remain identifiable throughout a person’s lifetime? Before adoption of any mainstream BMI-biometric innovations, it may be wise to ensure the security of BMI biometrics is as strong as the initial research suggests.

1. BMI Biometrics: What do we know?

 

2. BMI Biometrics: What don’t we know?

 

3. Security Applications: Beyond Biometrics

4. Alternative Modalities in BMI Biometrics

 

  1.       http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4445586/
  2.       http://www.nature.com/nmeth/journal/v11/n6/full/nmeth.2936.html
  3.       http://www.ajol.info/index.php/amhsr/article/view/112158
  4.       https://www.cert.org/insider-threat/best-practices/
  5.       D&d – Need to find this citation still again
  6.       – Lost this citation too
  7.       http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=812278
  8.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=813403
  9.       http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=4107575
  10.   https://www.researchgate.net/profile/B_Kumar8/publication/224194495_Subject_identification_from_electroencephalogram_(EEG)_signals_during_imagined_speech/links/54a97f8a0cf2eecc56e6c45f.pdf
  11.  http://www.theregister.co.uk/2014/12/29/german_minister_fingered_as_hackers_steal_her_thumbprint_from_a_photo/
  12.  http://www.sciencedirect.com/science/article/pii/S0921889014002632
  13.   S. Marcel, J. R. Millan. Person authentication using brainwaves (EEG) and maximum a posteriori model adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 29(4):743–748, April 2007.
  14.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6199830&tag=1
  15.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7126357&tag=1
  16.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7314193
  17.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6994220
  18.  https://kar.kent.ac.uk/49535/1/PIN_ijcb.pdf http://www.inderscienceonline.com/doi/abs/10.1504/IJBM.2014.060960
  19.   http://www.researchgate.net/profile/Patrizio_Campisi/publication/282122165_EEG_Biometrics_for_User_Recognition_using_Visually_Evoked_Potentials/links/56039ec808ae08d4f171779c.pdf
  20.   http://www.sciencedirect.com/science/article/pii/S0925231215004725
  21.   http://link.springer.com/chapter/10.1007/978-3-319-19713-5_13#page-1
  22.  http://download.springer.com/static/pdf/440/chp%253A10.1007%252F978-3-642-27733-7_9145-2.pdf?originUrl=http%3A%2F%2Flink.springer.com%2Freferenceworkentry%2F10.1007%2F978-3-642-27733-7_9145-2&token2=exp=1447987756~acl=%2Fstatic%2Fpdf%2F440%2Fchp%25253A10.1007%25252F978-3-642-27733-7_9145-2.pdf%3ForiginUrl%3Dhttp%253A%252F%252Flink.springer.com%252Freferenceworkentry%252F10.1007%252F978-3-642-27733-7_9145-2*~hmac=be7c889f4ec177bd4096057e765bb2169f3614dfd4323d2e28787a2e3b92393c
  23.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
  24.  http://www.ncbi.nlm.nih.gov/pubmed/19163618   

http://www.igi-global.com.proxy.libraries.rutgers.edu/gateway/chapter/full-text-pdf/7471

 

  1.   http://www.metaljournal.com.ua/assets/Journal/MMI-6/060-Jianfeng-Hu.pdf
  2.   http://csee.essex.ac.uk/staff/palaniappan/1930378.pdf
  3.   http://ieeexplore.ieee.org.proxy.libraries.rutgers.edu/stamp/stamp.jsp?tp=&arnumber=5634487
  4.   – Need to find citation for this one as well  
  5.   Auditory potentials – https://etd.ohiolink.edu/!etd.send_file?accession=ucin1439300974&disposition=inline
  6.   http://ojs.bibsys.no/index.php/NIK/article/view/243
  7.   http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7280344
  8.   http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7318985
    1.   Facial movements, eeg authentication using artifacts http://link.springer.com/chapter/10.1007/978-3-319-07995-0_34#page-1
    2.   Multi-level approach based on eye-blinking http://www.sciencedirect.com/science/article/pii/S0167865515002433
    3.   http://www.nature.com/neuro/journal/v18/n11/full/nn.4135.html
    4.   DOI is good  http://www.nature.com/nphoton/journal/v8/n6/full/nphoton.2014.107.html
    5.   DOI vs fMRI – http://www.ncbi.nlm.nih.gov/pubmed/23578579

    38.    http://ieeexplore.ieee.org/xpls/icp.jsp?arnumber=1619442

Benefits of EEG as biometric

What are the benefits of using EEG as the underpinning for a proof-of-individual network?

The use of biometrics, a pattern recognition system, for identification and verification of individuals, is now a widespread technology. With the shortcomings of classical biometrics like retinal scans and fingerprints among others, it has become necessary to come up with more innovative biometrics that are more secure. Bioelectric signals are an emerging candidate in this field. These are measurable electrical signals of low amplitude. Of interests among these are the electrocardiography and electroencephalography (Campisi, La Rocca & Scarano, 2012). This paper looks into the possible application of EEGs as a biometric and the benefits of its usage.

An EEG, or electroencephalogram, is a tracing of the electrical activity generated by the brain. Scientific research has shown that these wave patterns, produced by the electrical activity of neurons, are unique to individuals and can be used as a biometric (Knight, 2007). These waveforms are recorded through electrodes placed on the scalp. This confers the usage of EEG as a biometric its first advantage. The process of recording the electrical activity of the brain is non-invasive and thus a very safe procedure. Despite the cumbersome nature of the process, its non-invasive nature makes it easily applicable in person identification. In addition to the non-invasiveness of the procedure, the data collection process is largely simple. According to Palaniappan and Mandic (2007), once electrodes are placed correctly on the scalp, the waveforms are traced and the analysis is automated. The equipment is also portable making the process more appealing to use as a biometric.

 

Perhaps the greatest advantage of EEGs is its very confidential nature. Unlike biometrics like fingerprints and voice traits, an individual’s EEG corresponds to the individual’s mental tasks as dictated by neuronal connections and is very difficult to mimics. The EEG has been shown to be virtually impossible to copy. According to Paranjape et al., (2011) because it depends on the inner mental tasks unique to every individual, it cannot be reproduced by others. This makes the use of EEG effective in person identification. This security is increased by the fact that EEGs can be modified depending on the state of the subject (human detection). The resting EEG waveform is different from that obtained when the subject is mentally engaged and the pattern is different when eyes are open or closed and when open (Taguiam, 2017). This makes copying such information impossible. If this were not enough, the security is further boosted by the change in patterns when under stress. So, unlike fingerprints, retinal scans and voice patterns, an individual cannot be forced to reproduce their mental passphrase (Campisi, La Rocca & Scarano, 2012).

 

Finally, the use of bioelectrical signals for person identification can help in diagnosing certain abnormalities that may be indicated by an EEG during the person identification process. This confers an additional advantage to the field of medicine. Certain medical conditions including sleep disorders, convulsive disorders and lately tumors of the CNS may be diagnosed using EEGs (Paranjape et al., 2011). Use of EEGs might just increase the detection of such disorders based on incidental EEG findings.
In conclusion, while still under development, the use of EEG in biometric identification has myriads of potential benefits. High on this list is the very secure nature of EEGs as a biometric. Additionally, the collection of subject waveforms is non-invasive and very safe. The EEG machines are portable and their use is likely to increase early diagnosis of brain pathologies like convulsive disorders, sleep disorders, and even brain tumors.

 

References
Campisi, P., La Rocca, D., & Scarano, G. (2012). EEG for automatic person recognition. Computer, 45(7), 87-89.
Knight, W. (16th January, 2007). Brain Activity provides novel biometric key. New Scientist. Retrieved from https://www.newscientist.com/article/dn10963-brain-activity-provides-novel-biometric-key/
Palaniappan, R., & Mandic, D. P. (2007). EEG based biometric framework for automatic identity verification. The Journal of VLSI Signal Processing Systems for Signal, Image, and Video Technology, 49(2), 243-250.
Paranjape, R. B., Mahovsky, J., Benedicenti, L., & Koles, Z. (2011). The electroencephalogram as a biometric. In Electrical and Computer Engineering, 2001. Canadian Conference on (Vol. 2, pp. 1363-1366). IEEE.
Taguiam, R. A. (Jan 24th 2017). Brainwaves Cab Be Passwords, Scientists Explain How EEG Authentication Works. Nature World News. Retrieved from https://www.natureworldnews.com/articles/35169/20170124/brainwaves-passwords-scientists-explain-eeg-authentication-works.htm

Liveness-detectable biometrics, and cybersecurity

How can we use liveness-detectable biometrics to provide computational barriers to creating fake accounts and fake votes?

Network security today is more difficult than it needs to be, largely because it must be based on the assumption that a single user can hold multiple accounts or multiple IP addresses. The one commonality between e-mail spam, DDOS attacks, presidential election meddling, and blockchain fees is that they’re all based on a single person holding multiple accounts. This person may submit multiple forum posts, multiple http requests, or multiple high-fee cryptocurrency transactions for the sole purpose of attacking or clogging the network, ruining it for everyone else. You might know this just as “spam”, or as “fake news”.

If there were a walled garden of the internet where each user was guaranteed to have provably one account, based on their biometricity, just like in the real world. You could also ban users who spam the network with these inexpensive requests, and could hold each person’s posts to a higher degree of accountability (people wouldn’t submit fake news if they couldn’t shed their post’s biometric signature). In blockchain, the real reason blockchain payments aren’t cheaper than say Visa for millions of users is due to this fact; People are clogging the network with low-priority payments and spam payments.

hd1

So, building a biometric internet isn’t really that simple. You can’t just say “here’s my fingerprint, now create my account”, because a fingerprint is really just an image file. An attacker could copy the scan of your fingerprint and pretend to be you very easily, otherwise this would have been done before.

What you need is a signal that exhibits biometricity, unique features that identify you, but is also a liveness challenge, a puzzle that basically asks “are you human?” and “are you actually there, right now, submitting this biometric signal?”.  If what you submit equates to a yes, you can enter the garden. If you’ve ever had to type in numbers on a street sign, or crooked letters in all caps, or anything that’s like “I’m not a robot”, that’s a liveness challenge (Google’s reCAPTCHA). It’s a liveness challenge, but it doesn’t tie you to one account because it doesn’t exhibit biometricity. There’s nothing biological to distinguish your response from that over other users — it guarantees one person per computer, but it doesn’t guarantee you’re going to act honestly once you’re inside, because you’re still anonymous. A liveness challenge that also exhibited biometricity would ensure that whatever biometric signal you were submitting, like a fingerprint, wasn’t copy and pasted from a file, but rather the signal was generated quite recently, from an actual human being behind their computer screen.

A simple example of a challenge that exhibits both liveness detection and human detectability might be to have some sort of public ledger that is a string of random values, like a blockchain. Each block contains a random hash value that you can use to derive 20 random words out of a list of 1000. These words are derived from the hash value, meaning there is a relation between the block’s hash and this new, presentable stimulus. You could then present these 20 random words to a user and tell them, “quick, say these 20 random words out loud!”. A person says the words and their voice, which is unique to them, is recorded and propagated across the network, where it is confirmed by nodes to be a biometric unique to them. It’s a liveness challenge because it’s difficult for a computer program to generate audio that (1) exhibited biometricity and (2) contained words seeded by a very recent block (i.e. the computational power to generate this may expensive if the seeding block was found quite recently). The fact that you were able to say the words so quickly would indicate that the signal wasn’t generated in advance, or before the random value was submitted, which is a sign that the words came from a live human. A person’s voice is a good (but not great) biometric, so it exhibits biometric features as well.

hd2

Everything discussed so far is great, except for the fact that the NSA and big tech companies have been collecting voice data for years and would be really good at generating a voice simulation to say those 20 words within 3 minutes (or three seconds). We learned from the Titanic that nothing is ever going to be 100% secure, so voice data definitely has a role in our network, but as a more future-proof liveness challenge/biometric, imagine now you’re inputting an electrical signal that was generated based on a block hash. The signal passes over a user’s skin near their left ear and comes out their right ear. The signal as it leaves the right ear still contains elements of the original signal, but it has been modulated based on the unique biological properties of your skin. We then digitize the analog signal so that it can be transmitted to a network of computers which analyze it.

hd3

The most efficient blockchain will be one where every single user has provably one account. Decentralized systems are plagued by spam, so if you can unclog the network from these spam attacks, you can make the fees cheaper for everyone. Right now credit card companies take 2% off every transaction, a financial network with 0% fees is obviously going to be preferable; Money will always flow to the space with highest liquidity (unless it’s prone to corruption like EOS). To build this one-person-one-vote network you need a biometric that exhibits liveness, and one method for doing that is described here, using the unique properties of brainwave biometricities.

A long-term validation metric can be used as an appendage to this, as pioneered by OpenMined.org. Basically, to validate that data isn’t faked, a machine learning algorithm (over a time-consuming process) determines which data improves its recognition rates and which do not. If it were possible to fake out a machine learning algorithm with bad data, then ML theorists would already be doing it, and nobody would be paying for data. Thus, the most stable foundation for a network like this ultimately (and ironically) relies on artificial intelligence (in the long-term / for cementing transactions) as well as fluid biometrics and a reputation system (in the short-term / for immediate payouts).

Finally, we provide a financial incentive for identifying fake accounts based on biometric signals by creating a challengers-verifiers market. People are motivated by financial reward to check each biometric signal submitted and verified by block producers. Block producers are expected to submit fake data 1% of the time and distribute new oblio as a reward. See our github for the latest spec on this component.

Keep an eye out for our next post, where we’ll be delving into the fully-blockchain-compatible algorithm that reached higher identification rates on brainwaves than any other study we’ve seen.