Image and graphics . Part I : 11th International Conference, ICIG 2021, Haikou, China, August 6-8, 2021, proceedings / / Yuxin Peng [and five others], editors

• Pubblicazione: Cham, Switzerland : , : Springer, , [2021]; ©2021
• Descrizione fisica: 1 online resource (830 pages)
• Disciplina: 621.367
• Soggetto topico: Image processing - Digital techniques
• Persona (resp. second.): PengYuxin
• Nota di bibliografia: Includes bibliographical references and index.
• Nota di contenuto: Intro -- Preface -- Organization -- Contents - Part I -- Contents - Part II -- Contents - Part III -- Object Detection and Recognition -- L2-CVAEGAN: Feature Aligned Generative Networks for Zero-Shot Learning -- 1 Introduction -- 2 Related Work -- 2.1 Zero-Shot Learning -- 2.2 Generative Models -- 3 Methodology -- 3.1 Setup -- 3.2 Model Overview -- 3.3 Our Proposed Method -- 3.4 Evaluation Protocol -- 4 Experiments -- 4.1 Settings -- 4.2 Comparing with Different Methods -- 4.3 Visualization and Analyzing -- 4.4 Generalized Zero-Shot Learning -- 4.5 Effect of Semantic Combination -- 5 Conclusion -- References -- HQ-Trans: A High-Quality Screening Based Image Translation Framework for Unsupervised Cross-Domain Pedestrian Detection -- 1 Introduction -- 2 Related Work -- 2.1 Image Translation -- 2.2 Object Detection -- 2.3 Unsupervised Pedestrian Detection -- 3 Method -- 3.1 The First Screening -- 3.2 Scene Translation -- 3.3 The Second Screening -- 3.4 Pedestrian Detector -- 4 Experiments -- 4.1 Experiment Settings -- 4.2 Experimental Results and Analysis -- 4.3 Comparisons with Other Works -- 5 Conclusion -- References -- FER-YOLO: Detection and Classification Based on Facial Expressions -- 1 Introduction -- 2 Methodology -- 2.1 Model Architecture -- 2.2 Channel Attention -- 2.3 Depth-Wise Separable Convolution -- 3 Experimental Results and Analysis -- 3.1 Implementation Details -- 3.2 Ablation Experiments -- 3.3 Comparisons with State-of the Art Methods -- 3.4 Real-Time Facial Expression Detection via Camera -- 4 Conclusion -- References -- MSC-Fuse: An Unsupervised Multi-scale Convolutional Fusion Framework for Infrared and Visible Image -- 1 Introduction -- 2 Proposed Fusion Method -- 2.1 Network Architecture -- 2.2 The Basic Convolution Blocks -- 2.3 Loss Function -- 3 Experimental Results -- 3.1 Experimental Environment and Parameter Setting. ; 3.2 Ablation Study -- 3.3 Compared with Other Methods -- 4 Conclusions -- References -- Relation-Aware Reasoning with Graph Convolutional Network -- 1 Introduction -- 2 Related Works -- 2.1 Knowledge Graph Based Methods -- 2.2 Relational Reasoning -- 3 Proposed Framework -- 3.1 Knowledge Graph Construction -- 3.2 GCN for Visual Reasoning -- 3.3 Implementation Details -- 4 Experiments -- 4.1 Datasets and Evaluation -- 4.2 Experimental Results -- 4.3 Ablation Study -- 5 Conclusion -- References -- Feature Separation GAN for Cross View Gait Recognition -- 1 Introduction -- 2 Related Work -- 2.1 Generative Adversarial Networks -- 2.2 Generative Method for Gait Recognition -- 2.3 Gait Energy Image -- 3 Proposed Method -- 3.1 Network Structure -- 3.2 Separation Features -- 3.3 Constraint Block -- 3.4 Generator Optimizing -- 4 Experiment Result -- 4.1 Data Set -- 4.2 Implementation Details -- 4.3 Experimental Result -- 5 Conclusion -- References -- Global Feature Polishing Network for Glass-Like Object Detection -- 1 Introduction -- 2 Related Work -- 3 Proposed Framework -- 3.1 Overview -- 3.2 Global Perception Module -- 3.3 Multi-scale Refinement Module -- 3.4 Global Feature Polishing Module -- 4 Experiments -- 4.1 Experimental Setup -- 4.2 Comparisons with SOTAs -- 4.3 Ablation Analysis -- 5 Conclusion -- References -- Imitating What You Need: An Adaptive Framework for Detector Distillation -- 1 Introduction -- 2 Related Works -- 2.1 Object Detection -- 2.2 Knowledge Distillation -- 3 Method -- 3.1 Network Overview -- 3.2 Backbone Features -- 3.3 Classification Head -- 3.4 Bounding Box Regression Head -- 3.5 Adaptive Distillation Decay -- 4 Experiments -- 4.1 Experimental Setup -- 4.2 Ablation Study -- 4.3 Experimental Results -- 4.4 Comparison with Previous Distillation Methods -- 5 Conclusion -- References. ; Fine-Grained Classification of Neutrophils with Hybrid Loss -- 1 Introduction -- 2 Related Work -- 2.1 Generic Object Detection -- 2.2 Fined-Grained Visual Classification -- 2.3 Face Recognition -- 3 Methods -- 3.1 Single Cell Detection -- 3.2 EfficientNet Backbone -- 3.3 Loss Functions -- 4 Experiments -- 4.1 Implementation Details -- 4.2 Ablation Study -- 4.3 Evaluation Results -- 5 Conclusions -- References -- Open-Set Product Authentication Based on Deep Texture Verification -- 1 Introduction -- 2 The Proposed Texture Authentication Framework -- 3 Datasets and Experimental Results -- 3.1 Texture Datasets -- 3.2 Experimental Results -- 4 Conclusion -- References -- Moving Object Detection Based on Self-adaptive Contour Extraction -- 1 Introduction -- 2 Methods -- 2.1 Framework Description -- 2.2 Contour Extraction -- 2.3 Region Proposal Network -- 3 Simulation Experiments -- 3.1 Dataset and Environmental Settings -- 3.2 Results and Analysis -- 4 Conclusion -- References -- Small Infrared Aerial Target Detection Using Spatial and Temporal Cues -- 1 Introduction -- 2 Related Works -- 2.1 Single Frame Based Methods -- 2.2 Multiple Frames Abased Methods -- 3 Small Infrared Aerial Target Detection using Spatial and Temporal Cues -- 3.1 Target Candidate Detection for Each Frame -- 3.2 Target Detection Using Trajectory Constraints -- 4 Experiments -- 4.1 Experimental Settings -- 4.2 Target Detection from Image Sequence -- 5 Conclusions -- References -- Boundary Information Aggregation and Adaptive Keypoint Combination Enhanced Object Detection -- 1 Introduction -- 2 Related Works -- 3 Methods -- 3.1 Overview of BANet -- 3.2 Boundary Information Aggregation -- 3.3 Adaptive Keypoint Combination -- 4 Experiments -- 4.1 Comparisons with Other One-Stage Object Detectors -- 4.2 Ablation Study and Error Analyse -- 4.3 Visualization of Detection Results. ; 5 Conclusions -- References -- Accurate Oriented Instance Segmentation in Aerial Images -- 1 Introduction -- 2 Relate Work -- 2.1 Object Detection -- 2.2 Instance Segmentation -- 3 Method -- 3.1 Detection Network -- 3.2 Segmentation Network -- 4 Experiments -- 4.1 Datasets -- 4.2 Implementation Details -- 4.3 Ablation Study -- 4.4 Comparison with State-of-the-Art Methods -- 5 Conclusion -- References -- Six-Channel Image Representation for Cross-Domain Object Detection -- 1 Introduction -- 2 Related Work -- 3 Proposed Approach -- 3.1 Image-to-Image Translation -- 3.2 CNN Models -- 4 Experiments -- 4.1 Datasets -- 4.2 Experimental Evaluations -- 4.3 Experimental Results -- 5 Conclusion -- References -- Skeleton-Aware Network for Aircraft Landmark Detection -- 1 Introduction -- 2 Skeleton-Aware Landmark Detection Network -- 2.1 Global Geometrical Structure -- 2.2 Local Landmark Localization -- 2.3 Effectiveness of the Global Skeleton Structure -- 2.4 Implementation Detail -- 3 Experiments -- 4 Conclusions -- References -- Efficient Spectral Pyramid and Spectral-Spatial Feature Interactive Hyperspectral Image Classification -- 1 Introduction -- 2 Related Works -- 3 Proposed Network -- 3.1 Effective Spatial Extraction Based Interacted Network -- 3.2 Efficient Spectral Pyramid Module -- 3.3 Spectral-Spatial Feature Interaction Module -- 4 Experiment Setup -- 4.1 Experimental Datasets -- 4.2 Competing Methods -- 4.3 Parameter Setting -- 5 Classification Result -- 6 Ablation Analysis -- 6.1 Impact of the ESP Module -- 6.2 Impact of the Improved SSI Module -- 7 Conclusions -- References -- Semi-supervised Cloud Edge Collaborative Power Transmission Line Insulator Anomaly Detection Framework -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Insulator Detection on the Edge Device -- 3.2 Defect Recognition on the Cloud Server -- 4 Experiments. ; 4.1 Implementation Detail -- 4.2 Experiment Results -- 5 Conclusion -- References -- Illumination-Enhanced Crowd Counting Based on IC-Net in Low Lighting Conditions -- 1 Introduction -- 2 Related Work -- 2.1 Crowd Counting Based on Deep Learning -- 2.2 Crowd Counting Methods for Low Lighting Scenes -- 3 The Proposed IC-Net for Crowd Counting -- 3.1 Illumination Fusion Module (IFM) -- 3.2 Feature Cascading Module (FCM) -- 4 Experiments -- 4.1 Evaluation Metrics -- 4.2 Models' Performances with the Illumination Enhanced Image -- 4.3 Ablation Studies -- 5 Conclusion -- References -- An Open-Source Library of 2D-GMM-HMM Based on Kaldi Toolkit and Its Application to Handwritten Chinese Character Recognition -- 1 Introduction -- 2 Implement of 2D-HMM Based on Kaldi -- 2.1 Mathematical Formulation -- 2.2 Feature Extraction -- 2.3 Training Process -- 2.4 Testing Process -- 2.5 Implementation Details with Kaldi -- 3 Experiments -- 3.1 Task Description -- 3.2 Experimental Setup -- 3.3 Result Analysis -- 4 Conclusion -- References -- Automatic Leaf Diseases Detection System Based on Multi-stage Recognition -- 1 Introduction -- 2 Related Work -- 3 Network Construction Scheme:YOLOv4-Tiny + SEI-ResNet -- 3.1 YOLOv4-Tiny: A Fast Method to Remove Redundant Background -- 3.2 SEI-ResNet -- 3.3 Network Training -- 4 Experimental Results -- 4.1 Automatic Leaf Diseases Detection(ALDD) Network Positioning Test -- 4.2 ALDD Network Ablation Test -- 4.3 ALDD Network Classification Test -- 4.4 ALDD System Test -- 5 Conclusion -- References -- Aerial Image Object Detection Based on Superpixel-Related Patch -- 1 Introduction -- 2 Related Work -- 2.1 Superpixel Segmentation -- 2.2 Oriented Object Detection -- 3 Method -- 3.1 Cutting Aerial Image into Image Patches Based on Superpixel -- 3.2 Oriented Object Detection Based on YOLOv5 -- 4 Experiments -- 4.1 Dateset. ; 4.2 Superpixel Segmentation Comparison Experiments.
• Titolo autorizzato: Image and Graphics
• ISBN: 3-030-87355-2
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910502617903321
• Lo trovi qui: Univ. Federico II
• Serie: LNCS sublibrary. : SL 6, . -Image processing, computer vision, pattern recognition, and graphics ; ; Volume 12888.