Pattern Recognition. ICPR International Workshops and Challenges

Name: Pattern Recognition. ICPR International Workshops and Challenges | Virtual Event, January 10-15, 2021, Proceedings, Part V
Brand: Springer
Price: 96.29 EUR
Availability: OnlineOnly

Virtual Event, January 10-15, 2021, Proceedings, Part V

Alberto Del Bimbo Rita Cucchiara Stan Sclaroff Giovanni Maria Farinella Tao Mei Marco Bertini Hugo Jair Escalante Roberto Vezzani(Herausgeber*in)

Springer (Verlag)

Erschienen am 20. Februar 2021

XIX, 692 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-030-68821-9 (ISBN)

96,29 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
Foreword by General Chairs
Preface
Challenges
ICPR Organization
Contents - Part V
IMTA VII - Workshop on Image Mining Theory and Applications
IMTA-VII-2020: Workshop on Image Mining Theory and Applications
Workshop Description
Organization
IMTA Chairs
Scientific Committee
The Study of Improving the Accuracy of Convolutional Neural Networks in Face Recognition Tasks
1 Introduction
2 Data Preparation and Problem Statement
3 Comparison of Learning Processes and the Accuracy of Different Learning Approaches
4 Conclusion
References
Estimate of the Neural Network Dimension Using Algebraic Topology and Lie Theory
1 Motivation
2 Related Work
2.1 Differential Geometry in Neural Networks
2.2 Embedding Dimension of Neural Networks
3 The Manifold Assumptions
3.1 Smooth Manifolds
3.2 Lie Groups
4 Persistent Homology
4.1 Simplices
4.2 Simplicial Complexes
4.3 Associated Abelian Groups
4.4 Homological Persistence
4.5 Persistence Landscapes
5 Neural Networks
6 Counting Betti Numbers
7 Experimental Setting
7.1 Persistence Landscapes Hyperparameters
7.2 Neural Network Hyperparameters
7.3 Invertible Neural Network Architecture
7.4 Experimental Results
8 Conclusion
References
On the Depth of Gestalt Hierarchies in Common Imagery
1 Introduction
1.1 Related Work on Perceptual Grouping
1.2 Related Work on Aggregation of Parts
2 Gestalt Algebra
2.1 Operations
2.2 Search
3 Application to Example Imagery
3.1 Bottom-Up Analysis of Sections of the Image
3.2 Performance Assessment of the Application to a Set of Image Sections
3.3 Larger-Scale Gestalten on Global Image Level
3.4 Top-Down Constraint Enforcement
4 Discussion and Conclusion
References
Image Recognition Algorithms Based on the Representation of Classes by Convex Hulls
1 Introduction
2 The Definition of a Classifier Based on the Convex Hull Representation of Classes
3 Determining the Distance from the Test Point to the Convex Hull
3.1 Lite Distance Determination Method
3.2 A Method Based on Linear Programming
4 The Nearest Convex Hull Algorithm Based on the Linear Programming Method
5 Experimental Research
6 Conclusion
References
Tire Surface Segmentation in Infrared Imaging with Convolutional Neural Networks
1 Introduction
2 Methodology
2.1 Building a Dataset with a Heuristic Method
2.2 Proposed Architecture
2.3 Validation and Optimization
3 Experiments and Results
4 Conclusions
References
Human Action Recognition Using Recurrent Bag-of-Features Pooling
1 Introduction
2 Proposed Method
2.1 Bag-of-Features
2.2 Proposed Recurrent BoF
3 Experimental Evaluation
4 Discussion
5 Conclusions
References
Algorithms Based on Maximization of the Mutual Information for Measuring Parameters of Canvas Texture from Images
1 Introduction
2 Algorithms for Counting Threads
2.1 Filtering
2.2 Thresholding
2.3 Thread Counting
3 Experiment
4 Application of the Proposed Algorithms to Canvas Images of Paintings by F.S. Rokotov
5 Conclusions
References
Machine Learning Approach for Contactless Photoplethysmographic Measurement Verification
1 Introduction
2 Materials and Methods
2.1 Experimental Setup
2.2 Data Processing
3 Results
4 Conclusions
References
On New Kemeny's Medians
1 The Classic Kemeny's Median
2 The Kemeny's Metric Median
3 The Kemeny's Weighted Median
4 Weights Searching
5 Conclusion
References
Image Decomposition Based on Region-Constrained Smoothing
1 Introduction
2 Filtering Methods Preserving Image Structure
3 An Approach to Region-Constrained Smoothing
4 Similarity Factor
5 Filtering Based on Similarity Factor
6 Experiments
7 Conclusion
References
Machine Learning Based on Minimizing Robust Mean Estimates
1 Introduction
1.1 Outliers Problem
2 The Principle of Minimizing Robust Average Estimates
3 Outliers Insensitive Average Estimates
3.1 M-Averages
3.2 Smoothly Winsorized Averages
3.3 Smoothly Truncated Averages
4 Illustrative Examples
5 Conclusion
References
The Use of Machine Learning Methods to Detect Defects in Images of Metal Structures
1 Introduction
2 Features of Steel Structure Monitoring Tasks
3 Recognition of Structural Elements
4 Synthesis Detection of Defects in Steel Structures
5 Conclusions
References
Multiregion Multiscale Image Segmentation with Anisotropic Diffusion
1 Introduction
2 Anisotropic Diffusion Based Multiregion Multiscale Image Segmentation
2.1 Anisotropic Diffusion Filters
2.2 Intra-scale Splitting and Inter-scale Linking
2.3 Overall Scale Space Segmentation Method
3 Experimental Results
3.1 MMIS Segmentation Results
3.2 Medical Image Segmentation and Comparisons
4 Conclusions
References
The Test of Covariation Functions of Cylindrical and Circular Images
1 Introduction
2 Models of Homogeneous Images on a Cylinder and a Circle
3 Models of Inhomogeneous Images
4 The Test of Covariation Functions of Cylindrical and Circular Images
5 Conclusions
References
One-Class Classification Criterion Robust to Anomalies in Training Dataset
1 Introduction
2 Support Vector Data Description
3 Modified One-Class Classification Criterion Robust to the Presence of Outliers
4 The Differentiable Form of Criterion
5 Experimental Studies
6 Conclusion
References
Recognition of Tomographic Images in the Diagnosis of Stroke
1 Introduction
2 Related Work
3 Material and Method
3.1 Data Expansion
3.2 Convolutional Neural Network Architecture
3.3 Model Prediction Processing
4 Results of Experiments
5 Conclusion
References
Two-Stage Classification Model for Feather Images Identification
1 Introduction
2 Related Work
3 Dataset Construction
4 Experiment
4.1 Localization
4.2 Classes Normalization
4.3 Classification
4.4 Results
5 Conclusion
References
An Objective Comparison of Ridge/Valley Detectors by Image Filtering
1 Introduction
2 Ridge Extraction in Images
2.1 Hessian Matrix
2.2 Weingarten
2.3 Low Pass Filters for Ridge Detection
2.4 Oriented Filters
3 Experimental Results and Evaluation
3.1 Error Quantification and Evaluation Procedure
3.2 Synthetic Images Corrupted by Poisson Noise
3.3 Synthetic Images Corrupted by Speckle Noise
3.4 Synthetic Images Corrupted by Gaussian Noise
3.5 Evaluation with Real Images
4 Conclusion
References
High-Performance Algorithms Application for Retinal Image Segmentation Based on Texture Features
1 Introduction
2 An Algorithm for Separating a Laser Exposure Zone in the Retinal Image
3 Development of a High-Performance Segmentation Algorithm Based on the Use of Texture Features
3.1 An Algorithm for Histogram Generation
3.2 An Algorithm for Adjacency Matrix Generation
3.3 An Algorithm for Computing a Gradient Field
3.4 Algorithms for Computing Sum Vectors
4 Computing Texture Features
5 Experimental Study of the High-Performance Algorithm Acceleration
6 Conclusion
References
Learning Topology: Bridging Computational Topology and Machine Learning
1 Introduction
2 Persistence Homology: History and Basic Notions
2.1 Basics
2.2 History
2.3 Persistence Diagrams
3 PH-Based Descriptors and Implementations
3.1 Implementations
4 A Plethora of Applications
5 New Trends: PH into ML
5.1 Topological Layers
6 Conclusions
References
Library of Sample Image Instances for the Cutting Path Problem
1 Introduction
2 General Definition and Statements
3 Formation Principles for Specific Groups (Classes) of Image Instances in the Library
4 Computing Experiments
5 Conclusion
References
Interest Points Detection Based on Sign Representations of Digital Images
1 Introduction
2 Reduced Sign Representations and Their Graphs
3 Interest Points as Concentrators of Paths
4 Results of Computer Experiments
5 Conclusion
References
Memory Consumption and Computation Efficiency Improvements of Viola-Jones Object Detection Method for UAVs
1 Introduction
2 Viola-Jones Object Detection Method
2.1 Edge Haar-Like Features
3 Experiments
3.1 An Inflatable Life Raft PSN-10
3.2 An Oil Storage Tank
4 Conclusion
References
Automation of the Detection of Pathological Changes in the Morphometric Characteristics of the Human Eye Fundus Based on the Data of Optical Coherence Tomography Angiography
1 Introduction
2 Problem Statement
3 A Method for Automating the Detection of Pathological Changes in the Morphometric Characteristics of the Fundus
4 Experimental Investigation of the Proposed Method
5 Conclusions
References
MobileEmotiFace: Efficient Facial Image Representations in Video-Based Emotion Recognition on Mobile Devices
1 Introduction
2 Proposed Approach
3 Experimental Results
4 Conclusion
References
Basic Models of Descriptive Image Analysis
1 Introduction
2 Descriptive Image Models and Representations in Image Analysis
3 Models Reflecting the General Properties of the Image Recognition and Analysis Process
3.1 A Model Based on a Reversible Algebraic Closure
3.2 A Model Taking into Account the Equivalence Property of Images of the Same Class
3.3 A Model Based on Multiple Image Models and Multiple Classifiers
4 Models Characterizing the Architecture and Structure of the Recognition Process
4.1 Multilevel Model for Algorithms and Initial Data Combining in Image Recognition
4.2 Information Structure for Generating Descriptive Algorithmic Schemes for Image Recognition
5 Conclusion
References
Evaluation of Spectral Similarity Measures and Dimensionality Reduction Techniques for Hyperspectral Images
1 Introduction
2 Methods
2.1 Dimensionality Reduction Techniques Used in the Study
3 Experiments
4 Conclusion
References
Maximum Similarity Method for Image Mining
1 Introduction
2 Characterization of the Identification Problem in the Context of Image Mining
3 Image Representation by a Random Counts Set
4 Maximum Similarity Method
5 Conclusions
References
First Step Towards Creating a Software Package for Detecting the Dangerous States During Driver Eye Monitoring
1 Introduction
2 Algorithms for Eye Detection and Systems for Driver Monitoring
3 Hardware Analysis for Driver's Eyes Monitoring System
4 Collecting and Preparing Data for Training
5 Results of the Software Implementation
6 Conclusion
References
IUC 2020 - The 1st International Workshop on Human and Vehicle Analysis for Intelligent Urban Computing
Workshop on Human and Vehicle Analysis for Intelligent Urban Computing (IUC)
Workshop Description
Organization
Organization Chairs
Program Committee
Unbalanced Optimal Transport in Multi-camera Tracking Applications
1 Introduction
2 Related Works
2.1 Single View Multi-object Tracking
2.2 Multi-view Multi-object Tracking
3 Proposed Method
3.1 Targets Association Across Cameras as an Unbalanced Optimal Transport Problem
3.2 Ground Cost Learning for UOT-Based Targets Association Across Cameras
4 Experimental Results
4.1 Implementations
4.2 Benchmarking Performance
4.3 Performance Analysis
5 Conclusion
References
Arithmetic Evaluation System Based on MixNet-YOLOv3 and CRNN Neural Networks
1 Introduction
2 Related Work
3 Overall Architecture
3.1 Detection Branch
3.2 Recognition Branch
4 Experiments
4.1 Dataset
4.2 Evaluation
5 Conclusion
References
HSS-GCN: A Hierarchical Spatial Structural Graph Convolutional Network for Vehicle Re-identification
1 Introduction
2 Related Work
2.1 Vehicle Re-ID
2.2 GCN and GCN in Vehicle Re-ID
3 Methodology
3.1 Global Module
3.2 GCN Module with Hierarchical Spatial Graph
3.3 Joint Optimization of Multiple Losses
4 Experiments
4.1 Experiment Settings
4.2 Experimental Results and Analysis
4.3 Ablation Study
5 Conclusion
References
A Novel Multi-feature Skeleton Representation for 3D Action Recognition
1 Introduction
2 Related Work
3 The Proposed Method
3.1 Encoding Process
3.2 Important Angles Between Joints
4 Experiments and Analysis
4.1 Datasets
4.2 Implementation Details
4.3 Experimental Evaluation
5 Conclusions
References
R2SN: Refined Semantic Segmentation Network of City Remote Sensing Image
1 Introduction
1.1 Motivation
1.2 Contributions
2 Related Work
3 Our Approach
3.1 Structure
3.2 Architecture of Network
3.3 Focal Loss
3.4 Hinge Loss
4 Experiments
4.1 Implementation Details
4.2 Ablation Experiment
4.3 Comparison with Different Loss Functions
4.4 Comparison with State-of-the-Art Methods
5 Conclusion
References
Light-Weight Distilled HRNet for Facial Landmark Detection
1 Introduction
2 Proposed Method
2.1 Light-Weight HRNet
2.2 Dual Soft Argmax
2.3 Similarity-FeatureMap Knowledge Distillation
3 Experiment and Results
4 Conclusion
A Part Localization Results
References
DeepFM-Based Taxi Pick-Up Area Recommendation
1 Introduction
2 Related Work
3 Framework
4 Method
4.1 Score Prediction Based on DeepFM
4.2 Important Attribute of Area
4.3 Feature Extracting in DeepFM
4.4 Taxi Pick-Up Recommendation
5 Experiment
6 Summary
References
IWBDAF 2020 - International Workshop on Biometric Data Analysis and Forensics
International Workshop on Biometric Data Analysis and Forensics (IWBDA)
Organization
General Chairs
Program Committee
Blockchain-Based Iris Authentication in Order to Secure IoT Access and Digital Money Spending
1 Introduction
2 Blockchain
2.1 Transaction Validation
2.2 Blockchain Attacks
2.3 Blockchain and IoT
3 Related Work
4 Information Fusion of Prime Numbers and Biometrics for Encrypting Coding
5 A Smart Contract in Order to Prevent Double Spending
5.1 The Smart Contract
6 IIF Approach with a Novel or a Generalized Blockchain
6.1 Blockchain Enrollment and Wallet Creation
6.2 Permissioned Support Blockchain
6.3 Authentication
7 Analysis
7.1 Security of the Hybrid Key
7.2 Privacy
8 Conclusions and Future Work
References
DeepFakes Evolution: Analysis of Facial Regions and Fake Detection Performance
1 Introduction
2 Proposed Evaluation Framework
2.1 Facial Region Segmentation
2.2 Fake Detection Systems
3 Databases
3.1 UADFV
3.2 FaceForensics++
3.3 Celeb-DF
3.4 DFDC
4 Experimental Protocol
5 Experimental Results
5.1 Entire Face Analysis
5.2 Facial Regions Analysis
5.3 Comparison with the State of the Art
6 Conclusions
References
Large Scale Graph Based Network Forensics Analysis
1 Introduction
1.1 Organization of the Paper
2 Related Work
3 The Algorithm by Wang and Daniels
3.1 Global Reasoning
4 Our Algorithm
4.1 The MapReduce Paradigm
4.2 Apache Spark
4.3 Our MapReduce Version of the Algorithm by Wang and Daniels
5 Experimental Evaluation
5.1 Dataset
5.2 Hardware
5.3 Results
6 Conclusions
References
Deep Iris Compression
1 Introduction
2 Related Work
3 Deep Compression Model
4 Experimental Framework
5 Experiments and Analysis
6 Conclusion
References
IFEPE: On the Impact of Facial Expression in Head Pose Estimation
1 Introduction
1.1 Head Pose Estimation
1.2 Facial Expression Recognition
2 Methods and Materials
2.1 HP2IFS for Head Pose Estimation
2.2 FMPN-FER for Facial Expression Recognition
2.3 Datasets
3 Experiments and Results
4 Conclusions
References
Analysing and Exploiting Complexity Information in On-Line Signature Verification
1 Introduction
2 Proposed Approach
2.1 Signature Complexity Detector
2.2 On-Line Signature Verification System
3 Database
4 Experimental Protocol
5 Experimental Work
5.1 Exp. 1 - Analysis of Complexity for Baseline Systems
5.2 Exp. 2 - Training from Scratch 3 Systems, One System per Complexity
5.3 Exp. 3 - Training from Scratch and Changing the Baseline Architecture
5.4 Exp. 4 - Fine Tuning 3 Systems
5.5 Exp. 5 - Training a Global System with Balanced Classes
6 Conclusions and Future Work
References
A Novel Ensemble Framework for Face Search
1 Introduction
2 Related Works
2.1 Face Detection and Alignment
2.2 Face Feature Extraction
2.3 Approximate Nearest Neighbor Search (ANNS)
3 Proposed Framework
3.1 Face Detection and Alignment
3.2 Feature Extraction
3.3 Face Search
3.4 Fine Matching
4 Experiments and Results
5 Conclusion
References
Real-Time Thermal Face Identification System for Low Memory Vision Applications Using CNN
1 Introduction
2 Materials and Data
2.1 Thermal Camera
2.2 Microcontroller
2.3 Datasets
3 Related Works
4 Proposed CNN Architecture
5 Algorithm Pipeline
6 Experimental Analysis and Results
7 Conclusion
References
MADiMa 2020 - 6th International Workshop on Multimedia Assisted Dietary Management
Preface
Organization
Workshop Chairs
Publication Chair
Program Committee
Additional Reviewer
Assessing Individual Dietary Intake in Food Sharing Scenarios with Food and Human Pose Detection
1 Introduction
2 Related Work
3 Method
3.1 Dish Detection
3.2 Human Pose Estimation
3.3 Eating State Estimation
3.4 Bite Counting
4 Experiments
4.1 Dataset
4.2 Results
5 Conclusion
References
Recognition of Food-Texture Attributes Using an In-Ear Microphone
1 Introduction
2 Attribute Recognition Algorithms
2.1 Chew-Level Algorithm
2.2 Chewing-Bout-Level Algorithm
3 Data-Set and Evaluation Metrics
4 Evaluation and Results
5 Conclusions
References
Visual Aware Hierarchy Based Food Recognition
1 Introduction
2 Related Work
2.1 Image-Based Dietary Assessment
2.2 Food Detection and Food Localization
2.3 Food Image Recognition
2.4 Food Image Datasets
3 VIPER-FoodNet (VFN) Dataset
3.1 Food Categories
3.2 Semi-automatic Food Image Collection and Annotation
4 Food Recognition System
4.1 Food Localization
4.2 Food Classification
5 Experimental Results
5.1 Food Localization
5.2 Food Classification
5.3 Food Recognition
6 Discussion
7 Conclusion
References
Analysis of Chewing Signals Based on Chewing Detection Using Proximity Sensor for Diet Monitoring
1 Introduction
2 Hardware Design
3 Methodology
3.1 Data Collection
3.2 Chewing Detection
3.3 Chewing Analysis
4 Results
4.1 Chewing Detection
4.2 Chew Count Estimation
5 Discussions
6 Conclusion
References
Food Recognition in the Presence of Label Noise
1 Introduction
2 Related Work
2.1 Food Recognition
2.2 Noisy Labels
3 Method
3.1 Confusion Matrix (CM) Building
3.2 Noise Layer
4 Experimental Results
4.1 Dataset
4.2 Evaluation Metrics
4.3 Results
5 Conclusions
References
S2ML-TL Framework for Multi-label Food Recognition
1 Introduction
2 Related Works
3 Proposed Methodology
3.1 S2ML-TL Framework
3.2 Class Priors for Improved Transferability of Food Recognition
4 Validation
4.1 Datasets
4.2 Evaluation Metrics
4.3 Implementation Details
4.4 Results and Discussion
4.5 Ablation Study
5 Conclusions
References
UEC-FoodPix Complete: A Large-Scale Food Image Segmentation Dataset
1 Introduction
2 Related Work
3 Dataset Construction
4 Evaluation
5 Application 1: Region-Based Food Calorie Estimation
5.1 Method
5.2 Experimental Results
6 Application 2: Mask-Based Image Synthesis
7 Conclusions
References
Event Mining Driven Context-Aware Personal Food Preference Modelling
1 Introduction
2 Related Research
3 Food Event Model
3.1 The Causal Aspect
4 Food Preference Space: Taste Space
5 Experimental Design
6 Results
6.1 RQ1: Contextual Variation in Taste Profile
6.2 RQ2: Comparison of Prediction Performance
6.3 RQ3: Model Accuracy with Training Data Volume
7 Conclusion and Future Work
References
Analysis of Traditional Italian Food Recipes: Experiments and Results
1 Introduction
2 Related Works
3 Recipe Analysis
4 Recipe Visualization Application
5 CookIt Portal
6 Future Works and Conclusions
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken