SlideShare a Scribd company logo
1 of 1
Download to read offline
Towards Similarity Models in Police Photo
Lineup Assembling TasksLadislav Peška
Department of Software Engineering
Charles University in Prague, Czech Republic
ABSTRACT
Police photo lineups play a significant role
in the eyewitness identification process.
Unfortunately, there are many cases where
lineups have led to the incorrect
identification and conviction of innocent
suspects. One of the key factors affecting
the incorrect identification is the lack of
lineup fairness, i.e. that the suspect differs
significantly from other candidates.
Our current task to define a suitable inter-
person similarity metric and use it in a
lineup candidates recommender system.
We evaluated a combination of DCNN
descriptors of candidates’ photos and their
content-based features, which is further
tuned by the feedback of domain experts.
LinRec recommending algorithm further
considers the need for uniformity in
lineups. LinRec method was evaluated in a
realistic user study focused on lineup
fairness over solutions proposed by
domain experts.
IDENTIFICATION PROCESS and ERRORS IN PHOTO LINEUPS
RESULTS
• Precision (fraction of false identifications; the higher the better)
• Effective lineup size (ES, considering individual contribution of all
fillers; the higher the better)
• High variance in per-lineup results (easy vs. difficult suspects)
• No signifficance between mock witnesses’ age, education, gender,
or time to decision and correctness of identification
• Dataset size and variety is crucial (common problem in practice)
ONLINE EVALUATION: Mock witnesses + lineup fairness
• Dataset of over 4400 males, portrait photos, mostly Czech,
Vietnamese, Ukrainian and Slovak nationality
• Eight simulated suspects (4 Czech, 4 Vietnamese)
• Lineups assembled by the proposed method and 3 domain
experts, 6 persons in each lineup
• Short textual description of all 8 suspects
• Voluneers were asked to read a description of the suspect and
then try to identify him in the presented lineup
Should the lineup be completely fair, the success of mock witnesses
should be similar to a random guess
• In total 108 participants, mostly Czech
• Correct identification in 56% of cases
• Similar performance of proposed method with domain
experts => considerable reduction of manual labor
FUTURE WORK
- Co-operative dynamic lineups creation (recommendations + domain expert), demo application
- Local visual similarity (i.e. show only candidates with certain visual feature)
- Augumented rare features (i.e., hair style, tattoo) or substantial dataset extensions (legal complications)
- Explore boundaries of lineup fairness (e.g., are there too similar pairs of candidate-suspect?)
LinRec METHOD: Siamese network + enhanced uniformity
Hana Trojanová
Department of Psychology
Charles University in Prague, Czech Republic
1) Observe the event.
- numerous sources of error (poor lighting,
large distance, short time, fear, fatigue…)
- affects memory and recognition processes
- decrease witness certainty and reliability
2) Identify the offender
• decreased certainty may lead to incorrect
identification
• conviction of innocent persons
• Witnesses may follow the „best available choice“
selection
Real offender
might have not
been present in
the lineup
𝐿𝑂𝑆𝑆𝑆𝑖𝑎𝑚 =
𝑜 𝑖,𝑜 𝑗,𝑦 𝑖,𝑗)∈𝑇𝑠
IF 𝑦𝑖,𝑗 = 1: 𝑓𝑖 − 𝑓𝑗 2
2
ELSE: ma x( 𝐶 − 𝑓𝑖 − 𝑓𝑗 2
, 0
2
While assembling lineups, consider similarity to suspect as
well as similarities to already selected lineup fillers
ES = 𝐑 − 𝐑𝐑0
𝑐 𝑖∈𝐑
ma x( 0,
1
𝐑𝐑0
− 𝑓𝑖)
Lineup LinRec Expert 1 Expert 2 Expert 3
1 0.34 / 2.70 0.60 / 3.90 0.36 / 2.45 0.11 / 1.32
2 0.68 / 3.91 0.75 / 4.19 0.33 / 2.67 0.50 / 3.50
3 0.44 / 2.58 0.26 / 2.05 0.55 / 1.91 0.75 / 3.50
4 0.74 / 3.74 0.19 / 1.76 0.72 / 4.11 0.59 / 2.91
5 0.05 / 1.16 0.13 / 1.38 0.24 / 1.71 0.39 / 2.56
6 0.30 / 2.48 0.81 / 4.13 0.33 / 2.33 0.56 / 3.81
7 0.64 / 3.90 0.53 / 3.76 0.76 / 4.43 0.80 / 3.75
8 0.25 / 2.27 0.23 / 1.69 0.40 / 2.20 0.12 / 1.48
AVG(ES) 2.84 +/- 0.96 2.86 +/- 1.24 2.73 +/- 1.00 2.85 +/- 0.99
Per-lineup results: Precision / ES. Best results are bold, second best are underlined
Demo app: herkules.ms.mff.cuni.cz/lineit
Evaluation: herkules.ms.mff.cuni.cz/lineit-eval
Source codes & results: github.com/lpeska/LineRec
Contact: peska@ksi.mff.cuni.cz
This work was supported by grants GAUK-232217 and GACR-17-22224S

More Related Content

Similar to Towards Similarity Models in Police Photo Lineup Assembling Tasks

A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...
A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...
A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...IJERD Editor
 
Publication - The feasibility of gaze tracking for “mind reading” during search
Publication - The feasibility of gaze tracking for “mind reading” during searchPublication - The feasibility of gaze tracking for “mind reading” during search
Publication - The feasibility of gaze tracking for “mind reading” during searchA. LE
 
Criminals presentation
Criminals presentationCriminals presentation
Criminals presentationAnnye Braca
 
Canosa Saliency Based Decision Support
Canosa Saliency Based Decision SupportCanosa Saliency Based Decision Support
Canosa Saliency Based Decision SupportKalle
 
Pattern recognition and Machine Learning.
Pattern recognition and Machine Learning.Pattern recognition and Machine Learning.
Pattern recognition and Machine Learning.Rohit Kumar
 
Visual diagnostics for more effective machine learning
Visual diagnostics for more effective machine learningVisual diagnostics for more effective machine learning
Visual diagnostics for more effective machine learningBenjamin Bengfort
 
AIAA-SDM-PEMF-2013
AIAA-SDM-PEMF-2013AIAA-SDM-PEMF-2013
AIAA-SDM-PEMF-2013OptiModel
 
Face Detection and Recognition using Back Propagation Neural Network (BPNN)
Face Detection and Recognition using Back Propagation Neural Network (BPNN)Face Detection and Recognition using Back Propagation Neural Network (BPNN)
Face Detection and Recognition using Back Propagation Neural Network (BPNN)IRJET Journal
 
3 SAMPLING LATEST.pptx
3 SAMPLING LATEST.pptx3 SAMPLING LATEST.pptx
3 SAMPLING LATEST.pptxssuser504dda
 
Telefonica Lunch Seminar
Telefonica Lunch SeminarTelefonica Lunch Seminar
Telefonica Lunch SeminarNeal Lathia
 
Analysing a Complex Agent-Based Model Using Data-Mining Techniques
Analysing a Complex Agent-Based Model  Using Data-Mining TechniquesAnalysing a Complex Agent-Based Model  Using Data-Mining Techniques
Analysing a Complex Agent-Based Model Using Data-Mining TechniquesBruce Edmonds
 
PEMF2_SDM_2012_Ali
PEMF2_SDM_2012_AliPEMF2_SDM_2012_Ali
PEMF2_SDM_2012_AliMDO_Lab
 
Artificial Intelligence for Robotic AIFR
Artificial Intelligence for Robotic AIFRArtificial Intelligence for Robotic AIFR
Artificial Intelligence for Robotic AIFRadityabishts894
 
Detection and rectification of distorted fingerprints
Detection and rectification of distorted fingerprintsDetection and rectification of distorted fingerprints
Detection and rectification of distorted fingerprintsredpel dot com
 

Similar to Towards Similarity Models in Police Photo Lineup Assembling Tasks (20)

A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...
A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...
A Novel Framework For Numerical Character Recognition With Zoning Distance Fe...
 
Publication - The feasibility of gaze tracking for “mind reading” during search
Publication - The feasibility of gaze tracking for “mind reading” during searchPublication - The feasibility of gaze tracking for “mind reading” during search
Publication - The feasibility of gaze tracking for “mind reading” during search
 
Criminals presentation
Criminals presentationCriminals presentation
Criminals presentation
 
Canosa Saliency Based Decision Support
Canosa Saliency Based Decision SupportCanosa Saliency Based Decision Support
Canosa Saliency Based Decision Support
 
Pattern recognition and Machine Learning.
Pattern recognition and Machine Learning.Pattern recognition and Machine Learning.
Pattern recognition and Machine Learning.
 
Visual diagnostics for more effective machine learning
Visual diagnostics for more effective machine learningVisual diagnostics for more effective machine learning
Visual diagnostics for more effective machine learning
 
AIAA-SDM-PEMF-2013
AIAA-SDM-PEMF-2013AIAA-SDM-PEMF-2013
AIAA-SDM-PEMF-2013
 
Face Detection and Recognition using Back Propagation Neural Network (BPNN)
Face Detection and Recognition using Back Propagation Neural Network (BPNN)Face Detection and Recognition using Back Propagation Neural Network (BPNN)
Face Detection and Recognition using Back Propagation Neural Network (BPNN)
 
3 SAMPLING LATEST.pptx
3 SAMPLING LATEST.pptx3 SAMPLING LATEST.pptx
3 SAMPLING LATEST.pptx
 
Presentation of Visual Tracking
Presentation of Visual TrackingPresentation of Visual Tracking
Presentation of Visual Tracking
 
A0150106
A0150106A0150106
A0150106
 
A0150106
A0150106A0150106
A0150106
 
Telefonica Lunch Seminar
Telefonica Lunch SeminarTelefonica Lunch Seminar
Telefonica Lunch Seminar
 
Analysing a Complex Agent-Based Model Using Data-Mining Techniques
Analysing a Complex Agent-Based Model  Using Data-Mining TechniquesAnalysing a Complex Agent-Based Model  Using Data-Mining Techniques
Analysing a Complex Agent-Based Model Using Data-Mining Techniques
 
PEMF2_SDM_2012_Ali
PEMF2_SDM_2012_AliPEMF2_SDM_2012_Ali
PEMF2_SDM_2012_Ali
 
D04422730
D04422730D04422730
D04422730
 
SAC TRECK 2008
SAC TRECK 2008SAC TRECK 2008
SAC TRECK 2008
 
Artificial Intelligence for Robotic AIFR
Artificial Intelligence for Robotic AIFRArtificial Intelligence for Robotic AIFR
Artificial Intelligence for Robotic AIFR
 
Es1
Es1Es1
Es1
 
Detection and rectification of distorted fingerprints
Detection and rectification of distorted fingerprintsDetection and rectification of distorted fingerprints
Detection and rectification of distorted fingerprints
 

More from Ladislav Peska

Fuzzy D’Hondt’s Algorithm for On-line Recommendations Aggregation
Fuzzy D’Hondt’s Algorithm for On-line Recommendations AggregationFuzzy D’Hondt’s Algorithm for On-line Recommendations Aggregation
Fuzzy D’Hondt’s Algorithm for On-line Recommendations AggregationLadislav Peska
 
LineIT: Similarity search and recommendations for photo lineup assembling
LineIT: Similarity search and recommendations for photo lineup assemblingLineIT: Similarity search and recommendations for photo lineup assembling
LineIT: Similarity search and recommendations for photo lineup assemblingLadislav Peska
 
Off-line vs. On-line Evaluation of Recommender Systems in Small E-commerce
Off-line vs. On-line Evaluation of Recommender Systems in Small E-commerceOff-line vs. On-line Evaluation of Recommender Systems in Small E-commerce
Off-line vs. On-line Evaluation of Recommender Systems in Small E-commerceLadislav Peska
 
Linking Content Information with Bayesian Personalized Ranking via Multiple C...
Linking Content Information with Bayesian Personalized Ranking via Multiple C...Linking Content Information with Bayesian Personalized Ranking via Multiple C...
Linking Content Information with Bayesian Personalized Ranking via Multiple C...Ladislav Peska
 
Towards Complex User Feedback and Presentation Context in Recommender Systems
Towards Complex User Feedback and Presentation Context in Recommender SystemsTowards Complex User Feedback and Presentation Context in Recommender Systems
Towards Complex User Feedback and Presentation Context in Recommender SystemsLadislav Peska
 
Using the Context of User Feedback in Recommender Systems
Using the Context of User Feedback in Recommender SystemsUsing the Context of User Feedback in Recommender Systems
Using the Context of User Feedback in Recommender SystemsLadislav Peska
 
How to Interpret Implicit User Feedback
How to Interpret Implicit User FeedbackHow to Interpret Implicit User Feedback
How to Interpret Implicit User FeedbackLadislav Peska
 
Using Implicit Preference Relations to Improve Content-based Recommendations,...
Using Implicit Preference Relations to Improve Content-based Recommendations,...Using Implicit Preference Relations to Improve Content-based Recommendations,...
Using Implicit Preference Relations to Improve Content-based Recommendations,...Ladislav Peska
 
RecSys Challenge 2014, SemWexMFF group
RecSys Challenge 2014, SemWexMFF groupRecSys Challenge 2014, SemWexMFF group
RecSys Challenge 2014, SemWexMFF groupLadislav Peska
 

More from Ladislav Peska (9)

Fuzzy D’Hondt’s Algorithm for On-line Recommendations Aggregation
Fuzzy D’Hondt’s Algorithm for On-line Recommendations AggregationFuzzy D’Hondt’s Algorithm for On-line Recommendations Aggregation
Fuzzy D’Hondt’s Algorithm for On-line Recommendations Aggregation
 
LineIT: Similarity search and recommendations for photo lineup assembling
LineIT: Similarity search and recommendations for photo lineup assemblingLineIT: Similarity search and recommendations for photo lineup assembling
LineIT: Similarity search and recommendations for photo lineup assembling
 
Off-line vs. On-line Evaluation of Recommender Systems in Small E-commerce
Off-line vs. On-line Evaluation of Recommender Systems in Small E-commerceOff-line vs. On-line Evaluation of Recommender Systems in Small E-commerce
Off-line vs. On-line Evaluation of Recommender Systems in Small E-commerce
 
Linking Content Information with Bayesian Personalized Ranking via Multiple C...
Linking Content Information with Bayesian Personalized Ranking via Multiple C...Linking Content Information with Bayesian Personalized Ranking via Multiple C...
Linking Content Information with Bayesian Personalized Ranking via Multiple C...
 
Towards Complex User Feedback and Presentation Context in Recommender Systems
Towards Complex User Feedback and Presentation Context in Recommender SystemsTowards Complex User Feedback and Presentation Context in Recommender Systems
Towards Complex User Feedback and Presentation Context in Recommender Systems
 
Using the Context of User Feedback in Recommender Systems
Using the Context of User Feedback in Recommender SystemsUsing the Context of User Feedback in Recommender Systems
Using the Context of User Feedback in Recommender Systems
 
How to Interpret Implicit User Feedback
How to Interpret Implicit User FeedbackHow to Interpret Implicit User Feedback
How to Interpret Implicit User Feedback
 
Using Implicit Preference Relations to Improve Content-based Recommendations,...
Using Implicit Preference Relations to Improve Content-based Recommendations,...Using Implicit Preference Relations to Improve Content-based Recommendations,...
Using Implicit Preference Relations to Improve Content-based Recommendations,...
 
RecSys Challenge 2014, SemWexMFF group
RecSys Challenge 2014, SemWexMFF groupRecSys Challenge 2014, SemWexMFF group
RecSys Challenge 2014, SemWexMFF group
 

Recently uploaded

April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysismanisha194592
 
Call Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service BangaloreCall Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangaloreamitlee9823
 
Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...amitlee9823
 
Discover Why Less is More in B2B Research
Discover Why Less is More in B2B ResearchDiscover Why Less is More in B2B Research
Discover Why Less is More in B2B Researchmichael115558
 
Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...amitlee9823
 
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...amitlee9823
 
FESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfFESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfMarinCaroMartnezBerg
 
Midocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxMidocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxolyaivanovalion
 
VidaXL dropshipping via API with DroFx.pptx
VidaXL dropshipping via API with DroFx.pptxVidaXL dropshipping via API with DroFx.pptx
VidaXL dropshipping via API with DroFx.pptxolyaivanovalion
 
Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...
Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...
Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...Valters Lauzums
 
ALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptxALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptxolyaivanovalion
 
Probability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter LessonsProbability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter LessonsJoseMangaJr1
 
BDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort Service
BDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort ServiceBDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort Service
BDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort ServiceDelhi Call girls
 
Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...
Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...
Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...amitlee9823
 
Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...
Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...
Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...amitlee9823
 
Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...amitlee9823
 

Recently uploaded (20)

April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysis
 
Call Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service BangaloreCall Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Begur Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
 
Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
 
Discover Why Less is More in B2B Research
Discover Why Less is More in B2B ResearchDiscover Why Less is More in B2B Research
Discover Why Less is More in B2B Research
 
Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: 🍓 7737669865 🍓 High Profile Model Escorts | Bangalore ...
 
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Jalahalli Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
 
FESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfFESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdf
 
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get CytotecAbortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
 
Midocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxMidocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFx
 
VidaXL dropshipping via API with DroFx.pptx
VidaXL dropshipping via API with DroFx.pptxVidaXL dropshipping via API with DroFx.pptx
VidaXL dropshipping via API with DroFx.pptx
 
Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...
Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...
Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...
 
CHEAP Call Girls in Saket (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Saket (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICECHEAP Call Girls in Saket (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Saket (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICE
 
ALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptxALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptx
 
Probability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter LessonsProbability Grade 10 Third Quarter Lessons
Probability Grade 10 Third Quarter Lessons
 
BDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort Service
BDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort ServiceBDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort Service
BDSM⚡Call Girls in Mandawali Delhi >༒8448380779 Escort Service
 
Predicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science ProjectPredicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science Project
 
Call Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts Service
Call Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts ServiceCall Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts Service
Call Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts Service
 
Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...
Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...
Escorts Service Kumaraswamy Layout ☎ 7737669865☎ Book Your One night Stand (B...
 
Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...
Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...
Call Girls Hsr Layout Just Call 👗 7737669865 👗 Top Class Call Girl Service Ba...
 
Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
Call Girls Bommasandra Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
 

Towards Similarity Models in Police Photo Lineup Assembling Tasks

  • 1. Towards Similarity Models in Police Photo Lineup Assembling TasksLadislav Peška Department of Software Engineering Charles University in Prague, Czech Republic ABSTRACT Police photo lineups play a significant role in the eyewitness identification process. Unfortunately, there are many cases where lineups have led to the incorrect identification and conviction of innocent suspects. One of the key factors affecting the incorrect identification is the lack of lineup fairness, i.e. that the suspect differs significantly from other candidates. Our current task to define a suitable inter- person similarity metric and use it in a lineup candidates recommender system. We evaluated a combination of DCNN descriptors of candidates’ photos and their content-based features, which is further tuned by the feedback of domain experts. LinRec recommending algorithm further considers the need for uniformity in lineups. LinRec method was evaluated in a realistic user study focused on lineup fairness over solutions proposed by domain experts. IDENTIFICATION PROCESS and ERRORS IN PHOTO LINEUPS RESULTS • Precision (fraction of false identifications; the higher the better) • Effective lineup size (ES, considering individual contribution of all fillers; the higher the better) • High variance in per-lineup results (easy vs. difficult suspects) • No signifficance between mock witnesses’ age, education, gender, or time to decision and correctness of identification • Dataset size and variety is crucial (common problem in practice) ONLINE EVALUATION: Mock witnesses + lineup fairness • Dataset of over 4400 males, portrait photos, mostly Czech, Vietnamese, Ukrainian and Slovak nationality • Eight simulated suspects (4 Czech, 4 Vietnamese) • Lineups assembled by the proposed method and 3 domain experts, 6 persons in each lineup • Short textual description of all 8 suspects • Voluneers were asked to read a description of the suspect and then try to identify him in the presented lineup Should the lineup be completely fair, the success of mock witnesses should be similar to a random guess • In total 108 participants, mostly Czech • Correct identification in 56% of cases • Similar performance of proposed method with domain experts => considerable reduction of manual labor FUTURE WORK - Co-operative dynamic lineups creation (recommendations + domain expert), demo application - Local visual similarity (i.e. show only candidates with certain visual feature) - Augumented rare features (i.e., hair style, tattoo) or substantial dataset extensions (legal complications) - Explore boundaries of lineup fairness (e.g., are there too similar pairs of candidate-suspect?) LinRec METHOD: Siamese network + enhanced uniformity Hana Trojanová Department of Psychology Charles University in Prague, Czech Republic 1) Observe the event. - numerous sources of error (poor lighting, large distance, short time, fear, fatigue…) - affects memory and recognition processes - decrease witness certainty and reliability 2) Identify the offender • decreased certainty may lead to incorrect identification • conviction of innocent persons • Witnesses may follow the „best available choice“ selection Real offender might have not been present in the lineup 𝐿𝑂𝑆𝑆𝑆𝑖𝑎𝑚 = 𝑜 𝑖,𝑜 𝑗,𝑦 𝑖,𝑗)∈𝑇𝑠 IF 𝑦𝑖,𝑗 = 1: 𝑓𝑖 − 𝑓𝑗 2 2 ELSE: ma x( 𝐶 − 𝑓𝑖 − 𝑓𝑗 2 , 0 2 While assembling lineups, consider similarity to suspect as well as similarities to already selected lineup fillers ES = 𝐑 − 𝐑𝐑0 𝑐 𝑖∈𝐑 ma x( 0, 1 𝐑𝐑0 − 𝑓𝑖) Lineup LinRec Expert 1 Expert 2 Expert 3 1 0.34 / 2.70 0.60 / 3.90 0.36 / 2.45 0.11 / 1.32 2 0.68 / 3.91 0.75 / 4.19 0.33 / 2.67 0.50 / 3.50 3 0.44 / 2.58 0.26 / 2.05 0.55 / 1.91 0.75 / 3.50 4 0.74 / 3.74 0.19 / 1.76 0.72 / 4.11 0.59 / 2.91 5 0.05 / 1.16 0.13 / 1.38 0.24 / 1.71 0.39 / 2.56 6 0.30 / 2.48 0.81 / 4.13 0.33 / 2.33 0.56 / 3.81 7 0.64 / 3.90 0.53 / 3.76 0.76 / 4.43 0.80 / 3.75 8 0.25 / 2.27 0.23 / 1.69 0.40 / 2.20 0.12 / 1.48 AVG(ES) 2.84 +/- 0.96 2.86 +/- 1.24 2.73 +/- 1.00 2.85 +/- 0.99 Per-lineup results: Precision / ES. Best results are bold, second best are underlined Demo app: herkules.ms.mff.cuni.cz/lineit Evaluation: herkules.ms.mff.cuni.cz/lineit-eval Source codes & results: github.com/lpeska/LineRec Contact: peska@ksi.mff.cuni.cz This work was supported by grants GAUK-232217 and GACR-17-22224S