Computer analysis of images and patterns 19th international conference, CAIP 2021, virtual event, September 28-30, 2021, proceedings . Part II. / / Nicolas Tsapatsoulis [and five others] editors

• Pubblicazione: Cham, Switzerland : , : Springer, , [2021]; ©2021
• Descrizione fisica: 1 online resource (457 pages)
• Disciplina: 621.367
• Soggetto topico: Image processing - Digital techniques; Computer vision; Pattern recognition systems
• Persona (resp. second.): TsapatsoulisNicolas <1969->
• Nota di contenuto: Intro -- Preface -- Organization -- Contents - Part II -- Contents - Part I -- Machine Learning -- Deep Learning Based Automated Vickers Hardness Measurement -- 1 Introduction and Related Work -- 2 Methodology -- 2.1 Indentation Segmentation Using Convolutional Neural Network -- 2.2 Edge Extraction and Initial Indention Vertex Position Estimation -- 2.3 Precision Improvement -- 3 Experimental Framework -- 4 Experiments and Results -- 5 Conclusion -- References -- ElasticHash: Semantic Image Similarity Search by Deep Hashing with Elasticsearch -- 1 Introduction -- 2 Related Work -- 3 ElasticHash -- 3.1 Deep Hashing Model -- 3.2 Integration into ES -- 4 Experimental Evaluation -- 4.1 Results -- 5 Conclusion -- References -- Land Use Change Detection Using Deep Siamese Neural Networks and Weakly Supervised Learning -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 3.1 Multi Filter Multi-scale Deep Convolutional Neural Network -- 3.2 Siamese Neural Network -- 3.3 Generation of Change Detection Maps -- 4 Experimental Setup -- 4.1 Datasets' Description -- 4.2 Model Adaptation and Parameter Setting -- 5 Results -- 5.1 Ablation Analysis of the Proposed Model -- 6 Conclusions -- References -- AMI-Class: Towards a Fully Automated Multi-view Image Classifier -- 1 Introduction -- 2 Proposed Methodology -- 2.1 Notations -- 2.2 MVE Component -- 2.3 Multi-view CASH Solver AMI -- 3 Experiments and Results -- 3.1 MVE Component Validation -- 3.2 CASH Solver Validation -- 4 Conclusion and Discussion -- References -- How Realistic Should Synthetic Images Be for Training Crowd Counting Models? -- 1 Introduction -- 2 Related Work -- 3 Motivations and Goal of This Work -- 4 Synthetic Image Generation -- 5 Experimental Evaluation -- 5.1 Data Sets -- 5.2 Experimental Set-Up -- 5.3 Results -- 6 Conclusions -- References. ; Unsupervised Recognition of the Logical Structure of Business Documents Based on Spatial Relationships -- 1 Introduction and Context -- 2 State of the Art -- 3 Proposal -- 4 Spatial Contexts -- 4.1 Metadata to Captions Spatial Context (MCSC) -- 4.2 Metadata to Metadata Spatial Context (MMSC) -- 5 Voting and Detection Stages -- 5.1 Voting by Using Metadata Position -- 5.2 Voting by Using the Metadata to Captions Spatial Context -- 5.3 Voting by Using the Metadata Format -- 5.4 Voting by Using the MMSC -- 5.5 Detection Stage -- 6 Results -- 7 Conclusion -- References -- Feature Extraction -- The Method for Adaptive Material Classification and Pseudo-Coloring of the Baggage X-Ray Images -- 1 Introduction -- 2 Related Work -- 3 Proposed Method -- 3.1 Method Overview -- 3.2 Algorithm -- 3.3 Pseudo-Coloring -- 3.4 Evaluation Metrics -- 4 Experimental Results -- 5 Conclusions and Future Work -- References -- Sampling of Non-flat Morphology for Grey Value Images -- 1 Introduction -- 2 Fundamental Morphological Operations -- 2.1 Morphological Notions for Binary Images -- 2.2 Morphological Notions for Grey Scale Images -- 3 New Extensions -- 3.1 Operating on Grey Scale Images in the Sampled Domain -- 4 Conclusion -- References -- A Multi-scale Line Feature Detection Using Second Order Semi-Gaussian Filters -- 1 Introduction -- 2 Multi-scale Ridge Extraction: Related Works -- 2.1 Isotropic Filters -- 2.2 Oriented Filters for Line Feature Detection -- 3 Second-Derivative of a Semi-Gaussian Filter (SDSG) -- 3.1 Concept of the SDSG -- 3.2 Scale Fusion of the SDSG -- 4 Experimental Results and Evaluation -- 5 Conclusion -- References -- Experimental Analysis of Appearance Maps as Descriptor Manifolds Approximations -- 1 Introduction -- 2 Methodology -- 2.1 Appearance Map -- 2.2 Global Descriptor -- 2.3 Metrics -- 2.4 Evaluated Estimation Approaches -- 3 Experiments. ; 3.1 Approximation Accuracy of the Descriptor Manifold -- 3.2 Proof of Concept: The Image Gradient and the Descriptor Variation -- 3.3 Use Case: Image Derivative-Based Indicator for Appearance Maps -- 4 Conclusion -- References -- Building Hierarchical Tree Representations Using Homological-Based Tools -- 1 Introduction -- 2 Related Work -- 3 Background -- 4 Generation of the CAdjF, and *-trees -- 5 Conclusions, Applications and Future Research -- References -- Face Verification in Practice: The Case of Greek Artist Leonidas Arniotis -- 1 Introduction -- 2 Problem Statement -- 3 Methodology -- 3.1 Transfer Learning -- 3.2 Facial Points Matching -- 4 Results -- 5 Conclusion -- A Python Code -- References -- Object Recognition -- Spatio-Temporal Object Detection from UAV On-Board Cameras -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Short-Term Feature Aggregation -- 3.2 Long-Term Feature Aggregation -- 4 Experimental Results -- 4.1 Datasets -- 4.2 Implementation Details -- 4.3 Results -- 5 Conclusions -- References -- Automatic Watermeter Reading in Presence of Highly Deformed Digits -- 1 Introduction -- 2 Construction of Our Data Set -- 3 Data Set Preparation -- 4 Deep Learning Schemes -- 4.1 The CRNN Approach -- 4.2 The FCSRN Approach -- 5 The Connectionist Temporal Classification -- 6 Results -- 7 Conclusion -- References -- HR-Crime: Human-Related Anomaly Detection in Surveillance Videos -- 1 Introduction -- 2 HR-Crime Dataset -- 2.1 HR-Crime Statistics -- 2.2 Feature Extraction -- 3 Experiments -- 3.1 Datasets -- 3.2 Results and Discussion -- 4 Conclusion -- References -- Sequence-Based Recognition of License Plates with Severe Out-of-Distribution Degradations -- 1 Introduction -- 2 Related Work -- 3 Methods -- 3.1 CRNN Architecture -- 3.2 Generation of Synthetic Data Sets -- 3.3 Training Set Degradations -- 3.4 Test Set Degradations. ; 4 Evaluation -- 4.1 Robustness Testing on Synthetic Out-of-Distribution Data -- 4.2 Performance on Real World Data -- 5 Discussion -- 6 Conclusion -- References -- Exploiting Spatio-Temporal Coherence for Video Object Detection in Robotics -- 1 Introduction -- 2 Method Overview -- 2.1 Motion-Guided Propagation Model -- 2.2 Correspondence Step and Region Proposal -- 2.3 Bayesian Filtering over Object Class Labels -- 3 Experiments -- 3.1 Experimental Setup -- 3.2 Evaluation Metrics -- 3.3 Implementation Details -- 3.4 Experimental Results -- 4 Conclusions and Future Work -- References -- Face and Gesture -- A Study of General Data Improvement for Large-Angle Head Pose Estimation -- 1 Introduction -- 2 Related Work -- 2.1 Datasets -- 2.2 Networks -- 2.3 Evaluation -- 3 Optimizing Data Distribution by Conversion of Rotation Orders -- 3.1 Disadvantages of Traditional Pitch-Yaw-Roll Order -- 3.2 Study on Different Rotation Orders -- 3.3 Results of Optimizing Data Distribution -- 4 Data Enhancement by Random Rotation -- 4.1 Data Missing for Rare Poses -- 4.2 Solution to the Head Pose of the Rotated Image -- 4.3 Results of Data Enhancement -- 5 Conclusion -- References -- Knight Tour Patterns: Novel Handcrafted Feature Descriptors for Facial Expression Recognition -- 1 Introduction -- 2 Related Work -- 3 Proposed Methodology -- 3.1 Knight's Tour -- 3.2 kTP Feature Extraction -- 3.3 KTP Feature Extraction -- 4 Experimental Analysis -- 5 Conclusion and Future Work -- References -- Exploiting Visual Context to Identify People in TV Programs -- 1 Introduction -- 2 Related Work -- 3 Dataset -- 3.1 Motivation -- 3.2 Dataset Structure -- 4 Visual Context Metric Learning -- 4.1 Triplet Formation -- 4.2 Model Learning -- 5 Experimental Evaluation -- 5.1 Visual Context Metric Evaluation -- 5.2 Evaluation on a Face Verification Task of Doppelgangers. ; 6 Qualitative Results -- 7 Conclusion -- References -- Foreground-Guided Facial Inpainting with Fidelity Preservation -- 1 Introduction -- 2 Related Work -- 2.1 Foreground Facial Inpainting Framework -- 3 Architecture -- 3.1 Loss Function -- 4 Training and Experiments -- 4.1 Datasets -- 4.2 Method Comparison -- 4.3 Implementation -- 5 Results and Discussion -- 5.1 Quantitative Results -- 5.2 Qualitative Results -- 5.3 Semantic Inpainting with Fidelity Preservation -- 6 Conclusion -- References -- Talking Detection in Collaborative Learning Environments -- 1 Introduction -- 2 Methodology -- 3 Results -- 3.1 Head Detection System Results -- 3.2 Head Video Region Classification Results -- 3.3 Talking Activity Detection System -- 4 Conclusion -- References -- Facial Recognition in Collaborative Learning Videos -- 1 Introduction -- 2 Methods -- 2.1 Computation of Face Prototypes -- 2.2 Video Faces Recognizer -- 3 Results -- 4 Conclusion -- References -- Guess the Age Contest -- Guess the Age 2021: Age Estimation from Facial Images with Deep Convolutional Neural Networks -- 1 Introduction -- 2 Contest Task -- 3 Methods -- 4 Evaluation Metrics -- 5 Results -- 6 Discussion -- 7 Conclusion -- References -- Real-Time Age Estimation from Facial Images Using YOLO and EfficientNet -- 1 Introduction -- 2 Related Work -- 3 Proposed Method -- 3.1 Face Detection -- 3.2 Age Estimation -- 4 Experiment -- 4.1 Datasets and Setting -- 4.2 Metrics -- 4.3 Results -- 5 Conclusion -- References -- Single View Facial Age Estimation Using Deep Learning with Cascaded Random Forests -- 1 Introduction -- 2 Deep Learning Methods for Age Estimation -- 2.1 Pre-processing -- 2.2 ResNeXt CNN -- 2.3 Two-Layer Random Forest (TLRF) -- 2.4 Training the Deep Architectures -- 3 Experimental Results -- 3.1 ResNeXt -- 3.2 Two-Layer Random Forest (TLRF). ; 3.3 Generalizability Performance Using the Withheld GTA Data.
• Titolo autorizzato: Computer Analysis of Images and Patterns
• ISBN: 3-030-89131-3
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 996464391503316
• Lo trovi qui: Univ. di Salerno
• Serie: Lecture notes in computer science ; ; 13053.