60 4. SEARCH PRUNING IN VIDEO SURVEILLANCE SYSTEMS4.8 SummaryThe current chapter provided statistical analysis <strong>of</strong> the gain and reliability in pruning the searchover large data sets, where these sets are random and where there is a possibility that the pruningmay entail errors. In this setting, pruning plays the role <strong>of</strong> pre-filtering, similar to techniques suchas video indexing. The analysis may <strong>of</strong>fer insight on better designing pre-filtering algorithms fordifferent search settings. We further studied nine different, actual, s<strong>of</strong>t biometric systems, as wellas analyzed and experimented with factors like average error, pruning gain and goodput. Usingthese factors, we provided a quantifiable comparison <strong>of</strong> these systems. Furthermore we identifiedrelations between SBS enhancement, error probabilityP err , pruning gain r and goodput U. Thesefindings bring to the fore some SBS design aspects. Finally we gave insight on the computationalcost reduction related to person recognition systems with a pruning mechanism. This insight revealedsome <strong>of</strong> the benefits <strong>of</strong> applying SBS for pre filtering.We here studied and analyzed the pertinent characteristics related to search pruning performedby SBSs. In the next chapter we examine a third scenario (after human identification and pruningthe search), namely human re-identification. In such a scenario in what follows, we explore thecapability and limitations <strong>of</strong> existing SB algorithms. We hereby introduce an additional challenge<strong>of</strong> frontal-to-side pose variation.
61Chapter 5Frontal-to-side person re–identificationTypically biometric face-recognition algorithms are developed, trained, tested and improvedunder the simplifying assumption <strong>of</strong> frontal-to-frontal person recognition. Such algorithms thoughare challenged when facing scenarios that deviate from the training setting, such as for examplein the presence <strong>of</strong> non-constant viewpoints, including the frontal-to-side scenario. Most personrecognition algorithms, whether holistic or based on facial features, only manage to optimally handlepose differences that are less than about15 degrees. As a result, a variation in the pose is <strong>of</strong>tena more dominant factor than a variation <strong>of</strong> subjects. This aspect <strong>of</strong> pose variation comes to the forein video surveillance, where a suspect may be pictured firstly frontal, whereas the correspondingtest images could be captured from the side, thus introducing a frontal-to-side recognition problem.Towards handling this problem, we draw as already in the chapters 3 and 4 from the wayhumans perform frontal-to-side recognition, that is by using simple and evident traits like hair,skin and clothes color. One <strong>of</strong> our tasks here is to get some insight into the significance <strong>of</strong>these traits, specifically the significance <strong>of</strong> using hair, skin and clothes patches for frontal-tosidere-identification. We mention that we work on the color FERET dataset [Fer11] with frontalgallery images for training, and side (pr<strong>of</strong>ile) probe images for testing. Towards achieving reidentification,the proposed algorithm first analyzes the color and texture <strong>of</strong> the three patches, aswell as their intensity correlations. This analysis is then followed by the construction <strong>of</strong> a single,stronger classifier that combines the above measures, to re-identify the person from his or herpr<strong>of</strong>ile.5.1 Related workPose invariant face recognition has been addressed in different approaches which, as describedin [PEWF08], can be classified in following three categories:– mapping methods: construction <strong>of</strong> a 3D model based on more than one 2D images (see [IHS05])– geometric methods: construction <strong>of</strong> a 3D model based on a single 2D image (see [SVRN07])– statistical methods: statistical learning methods that relate frontal to non-frontal poses (see [PEWF08].An overview <strong>of</strong> these frontal-to-side face recognition methods was given in [ZG09], which workalso addressed some <strong>of</strong> the methods’ limitations in handling different pose variations. Such methodscan be originally found in [WMR01] and [WAS + 05], which recorded a true recognition rate<strong>of</strong> 50-60% over an authentication group <strong>of</strong> 100 subjects. Better results on pose-variant face recognitionwere recorded in [PEWF08] which employed statistical methods to achieve reliability <strong>of</strong>92% over an authentication group with the same size.
- 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: 59for one person, for trait t, t =
- 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 154 and 155:
7- Not requiring the individual’s
- 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.