7– Not requiring the individual’s cooperation: Consent and contribution from the subject aregenerally not needed.– Preserving human privacy: The stored signatures are visually available to everyone andserve in this sense privacy.The plethora or utilities has motivated an increasing number <strong>of</strong> research activities related tos<strong>of</strong>t biometrics. We give an overview <strong>of</strong> scientific work gaining from the benefits related to s<strong>of</strong>tbiometrics.Related work We here outline work, pertinent to s<strong>of</strong>t biometrics. This overview does not claimto be an exhaustive state <strong>of</strong> the art, but rather a highlight selection on performed scientific studies.S<strong>of</strong>t biometrics is a relatively novel topic and related work enfolds over several research fields.Recent work can be mainly classified in three research fields:1. The first and largest field includes the study and identification <strong>of</strong> traits and associated imageprocessing algorithms for classification and detection <strong>of</strong> such.2. The second fast growing field identifies operational scenarios for the aforementioned algorithmsand provides experimental results for such scenarios.3. The third and smallest field comprises <strong>of</strong> the global and theoretical investigation <strong>of</strong> theemployment <strong>of</strong> s<strong>of</strong>t biometrics applications and related studies.Scientific works belonging to the first field cover algorithms for traits such as iris pattern, seein [SBS10], or facial marks, see in [JP09].The second field can be sub-classified in subgroups which differentiate the way s<strong>of</strong>t biometricsare employed, as stand–alone systems, as pre-filtering mechanisms <strong>of</strong> bigger systems, or as fusedparallel systems. Related scenarios include continuous authentication [NPJ10], video surveillancesee [DFBS09], [FDL + 10], [MKS10], person verification [ZESH04] and moreover person identification[PJ10]. An interesting recent associated scenario for SBS based person identification isthe recognition <strong>of</strong> faces in triage images <strong>of</strong> mass disaster situations [CO11].Finally the third field involves studies on the placement <strong>of</strong> s<strong>of</strong>t biometrics in applications suchas forensics [JKP11] and human metrology [ACPR10].Bag <strong>of</strong> facial s<strong>of</strong>t biometrics for human identificationWe consider the case where a SBS can distinguish between a set <strong>of</strong> traits (categories), whichset is large enough to allow for the classification that achieves human identification. The concept<strong>of</strong> person identification based on s<strong>of</strong>t biometrics originates in the way humans perform face recognition.Specifically human minds decompose and hierarchically structure complex problems int<strong>of</strong>ractions and those fractions into further sub-fractions, see [Ley96], [Sim96]. Consequently facerecognition performed by humans is the division <strong>of</strong> the face in parts, and subsequent classification<strong>of</strong> those parts into categories. Those categories can be naturally <strong>of</strong> physical, adhered or behavioralnature and their palette includes colors, shapes or measurements, what we refer to here as s<strong>of</strong>tbiometrics. The key is that each individual can be categorized in terms <strong>of</strong> such characteristics,by both humans or by image processing algorithms. Although features such as hair, eye and skincolor, facial hair and shape, or body height and weight, gait, cloth color and human metrologyare generally non distinctive, a cumulative combination <strong>of</strong> such features provides an increasinglyrefined and explicit description <strong>of</strong> a human. SBSs for person identification have several advantagesover classical biometric systems, as <strong>of</strong> non intrusiveness, computational and time efficiency,human compliance, flexibility in pose- and expression-variance and furthermore an enrolment freeacquirement in the absence <strong>of</strong> consent and cooperation <strong>of</strong> the observed person. S<strong>of</strong>t biometricsallow for a reduced complexity determination <strong>of</strong> an identity. At the same time though, the named
8reduced computational complexity comes with restrictions on the size <strong>of</strong> an authentication group.It becomes apparent that a measure <strong>of</strong> performance must go beyond the classical biometric equalerror rate <strong>of</strong> the employed detectors and include a different and new parametrization. Our generalinterest here is to provide insightful mathematical analysis <strong>of</strong> reliability <strong>of</strong> general s<strong>of</strong>t biometricsystems, as well as to concisely describe the asymptotic behavior <strong>of</strong> pertinent statistical parametersthat are identified to directly affect performance. Albeit its asymptotic and mathematicalnature, the approach aims to provide simple expressions that can yield insight into handling reallife surveillance systems.We introduce the setting <strong>of</strong> interest, which corresponds to the general scenario where, out <strong>of</strong> alarge population, an authentication group is randomly extracted as a random set <strong>of</strong> n people, out<strong>of</strong> which one person is picked for identification (and is different from all the other members <strong>of</strong>the authentication group). We note that this general scenario is consistent with both, the case <strong>of</strong>person verification as well as <strong>of</strong> identification. A general s<strong>of</strong>t-biometric system employs detectionthat relates toλs<strong>of</strong>t biometric traits (hair color, skin color, etc), where each traiti,i = 1,2,...,λ,is subdivided into µ i trait instances, i.e., each trait i can take one <strong>of</strong> µ i values. We henceforthdenote as category to be any λ-tuple <strong>of</strong> different trait-instances, and we let Φ = {φ i } ρ i=1 definea set <strong>of</strong> all ρ categories, i.e., the set <strong>of</strong> all ρ combinations <strong>of</strong> s<strong>of</strong>t-biometric trait-instances. Thenumber <strong>of</strong> ρ, that the system is endowed with, is given byρ = Π λ i=1µ i (1)In this setting we elaborate on pertinent factors, such as those <strong>of</strong> the authentication group,traits, traits instances, overall categories and their interrelations. We then proceed to introduce andexplain the event <strong>of</strong> collision, which is <strong>of</strong> significant character when employing SBSs for personidentification. event where any two or more subjects belong in the same category φ. Focusing ona specific subject, we say that this subject experiences interference if he/she belongs in a categorywhich also includes other subjects from the authentication group. In regards to this, we are interestedin gaining insight on two probability measures. The first measure is the probability p(n;ρ)that the authentication group <strong>of</strong> size n, chosen randomly from a large population <strong>of</strong> subjects, issuch that there exist two subjects within the group that collide. We briefly note the relationship <strong>of</strong>p(n;ρ) to the famous birthday paradox. For the other measure <strong>of</strong> system reliability, we considerthe case where an authentication group <strong>of</strong> size n is chosen randomly from a large population <strong>of</strong>subjects, and where a randomly chosen subject from within this authentication group, collideswith another member <strong>of</strong> the same group. We denote this probability asq(n), and note that clearlyq(n) < p(n). To clarify, p(n) describes the probability that interference exists, even though itmight not cause error, whereas q(n) describes the probability <strong>of</strong> an interference induced error.Example: In a group <strong>of</strong> N subjects p(n) would describe the probability that any two subjects willbelong to the same category φ x . On the other handq(n) reflects the probability that a specific subjectwill interfere with one or more <strong>of</strong> theN−1 remaining subjects. In the following we provide asimulation <strong>of</strong> the probability <strong>of</strong> identification error, in the setting <strong>of</strong> interest, under the assumptionthat the errors are due to interference, i.e., under the assumptions that errors only happen if andonly if the chosen subject shares the same category with another person from the randomly chosenauthentication group. This corresponds to the setting where the s<strong>of</strong>t-biometric approach cannotprovide conclusive identification. In the simulation, the larger population consisted <strong>of</strong> 646 peoplefrom the FERET database, and the simulation was run for different sizes n <strong>of</strong> the authenticationgroup. The probability <strong>of</strong> identification error is described in the following figure.As a measure <strong>of</strong> the importance <strong>of</strong> each trait, Figure 1 describes the collision probability whendifferent traits are removed. The presence <strong>of</strong> moustache and beard seem to have the least influenceon the detection results, whereas hair and eye color have the highest impact on distinctiveness.
- Page 1:
FACIAL SOFT BIOMETRICSMETHODS, APPL
- Page 5:
AcknowledgementsThis thesis would n
- Page 8:
6hair, skin and clothes. The propos
- Page 11 and 12:
97 Practical implementation of soft
- Page 13 and 14:
11Notations used in this workE : st
- Page 15 and 16:
13Chapter 1IntroductionTraditional
- Page 17 and 18:
15event of collision, which is of s
- Page 19 and 20:
17ric. In Section 6.6 we employ the
- Page 21 and 22:
19Chapter 2Soft biometrics: charact
- Page 23 and 24:
21is the fusion of soft biometrics
- Page 25 and 26:
23plied on low resolution grey scal
- Page 27 and 28:
25Chapter 3Bag of facial soft biome
- Page 29 and 30:
27In this setting we clearly assign
- Page 31 and 32:
29Table 3.1: SBSs with symmetric tr
- Page 33 and 34:
31corresponding to p(n,ρ). Towards
- Page 35 and 36:
the same category (all subjects in
- Page 37 and 38:
3.5.2 Analysis of interference patt
- Page 39 and 40:
an SBS by increasing ρ, then what
- Page 41 and 42:
39Table 3.4: Example for a heuristi
- Page 43 and 44:
41for a given randomly chosen authe
- Page 45 and 46:
43Chapter 4Search pruning in video
- Page 47 and 48:
45Figure 4.1: System overview.SBS m
- Page 49 and 50:
472.52rate of decay of P(τ)1.510.5
- Page 51 and 52:
49to be the probability that the al
- Page 53 and 54:
51The following lemma describes the
- Page 55 and 56:
534.5.1 Typical behavior: average g
- Page 57 and 58:
55n = 50 subjects, out of which we
- Page 59 and 60:
5710.950.9pruning Gain r(vt)0.850.8
- Page 61 and 62:
59for one person, for trait t, t =
- Page 63 and 64:
61Chapter 5Frontal-to-side person r
- Page 65 and 66:
63Figure 5.1: Frontal / gallery and
- Page 67 and 68:
6510.90.80.7Skin colorHair colorShi
- Page 69 and 70:
6710.90.80.70.6Perr0.50.40.30.20.10
- Page 71 and 72:
69Chapter 6Soft biometrics for quan
- Page 73 and 74:
71raphy considerations include [BSS
- Page 75 and 76:
73Figure 6.3: Example image of the
- Page 77 and 78:
75A direct way to find a relationsh
- Page 79 and 80:
77- Pearson’s correlation coeffic
- Page 81 and 82:
79shown to have a high impact on ou
- Page 83 and 84:
81Chapter 7Practical implementation
- Page 85 and 86:
834) Eye glasses detection: Towards
- Page 87 and 88:
857.2 Eye color as a soft biometric
- Page 89 and 90:
87Table 7.5: GMM eye color results
- Page 91 and 92:
89and office lights, daylight, flas
- Page 93 and 94:
917.5 SummaryThis chapter presented
- Page 95 and 96:
93Chapter 8User acceptance study re
- Page 97 and 98:
95Table 8.1: User experience on acc
- Page 99 and 100:
97scared of their PIN being spying.
- Page 101 and 102:
99Table 8.2: Comparison of existing
- Page 103 and 104: 101ConclusionsThis dissertation exp
- Page 105 and 106: 103Future WorkIt is becoming appare
- Page 107 and 108: 105Appendix AAppendix for Section 3
- Page 109 and 110: 107- We are now left withN −F = 2
- Page 111 and 112: 109Appendix BAppendix to Section 4B
- Page 113 and 114: 111Blue Green Brown BlackBlue 0.75
- Page 115 and 116: 113Appendix CAppendix for Section 6
- Page 117 and 118: 115Appendix DPublicationsThe featur
- Page 119 and 120: 117Bibliography[AAR04] S. Agarwal,
- Page 121 and 122: 119[FCB08] L. Franssen, J. E. Coppe
- Page 123 and 124: 121[Ley96] M. Leyton. The architect
- Page 125 and 126: 123[RN11] D. Reid and M. Nixon. Usi
- Page 127 and 128: 125[ZG09] X. Zhang and Y. Gao. Face
- Page 129: 2Rapporteurs:Prof. Dr. Abdenour HAD
- Page 132 and 133: Biométrie faciale douce 2Les terme
- Page 134 and 135: Biométrie faciale douce 4une perso
- Page 136 and 137: Couleur depeauCouleur descheveuxCou
- Page 138 and 139: Biométrie faciale douce 8Nous nous
- Page 140 and 141: Biométrie faciale douce 103. Proba
- Page 142 and 143: Biométrie faciale douce 12l’entr
- Page 144 and 145: Biométrie faciale douce 14Figure 6
- Page 146 and 147: Biométrie faciale douce 16pages 77
- Page 148 and 149: Reviewers:Prof. Dr. Abdenour HADID,
- Page 150 and 151: 3hair, skin and clothes. The propos
- Page 152 and 153: person in the red shirt”. Further
- Page 156 and 157: 9Probability of Collision10.90.80.7
- Page 158 and 159: 11the color FERET dataset [Fer11] w
- Page 160 and 161: 13Table 2: Table of Facial soft bio
- Page 162 and 163: 15Chapter 1PublicationsThe featured
- Page 164 and 165: 17Bibliography[ACPR10] D. Adjeroh,
- Page 166 and 167: 19[ZESH04] R. Zewail, A. Elsafi, M.