Computer vision and image processing : 5th international conference, CVIP 2020, Prayagraj, India, December 4-6, 2020, revised selected papers, part III / / Satish Kumar Singh [and three others] (editors)

• Pubblicazione: Singapore : , : Springer, , [2021]; ©2021
• Descrizione fisica: 1 online resource (556 pages)
• Disciplina: 006.37
• Soggetto topico: Computer vision; Image processing - Digital techniques
• Persona (resp. second.): SinghSatish Kumar
• Nota di contenuto: Intro -- Preface -- Organization -- Contents - Part III -- U-Net-Based Approach for Segmentation of Tables from Scanned Pages -- 1 Introduction -- 2 Related Work -- 3 Methods -- 4 Results -- 4.1 Observations -- 5 Conclusions -- References -- Air Writing: Tracking and Tracing -- Abstract -- 1 Introduction -- 2 Related Work -- 3 Proposed Method -- 3.1 Hand Detection with SSD -- 3.2 Noncontinuous Air Writing -- 3.3 Selective Erasing -- 4 Evaluation -- 5 Conclusion -- References -- Mars Surface Multi-decadal Change Detection Using ISRO's Mars Color Camera (MCC) and Viking Orbiter Images -- Abstract -- 1 Introduction -- 2 Data Processing Steps -- 2.1 Overlap Region Extraction -- 2.2 Geometric Transformation -- 2.3 SIFT Based Image Registration -- 2.4 MAD Based Change Detection -- 3 Comparison with Other Techniques -- 4 Processing Workflow Developed -- 5 Results Achieved -- 6 Conclusion -- Acknowledgement -- References -- Deep over and Under Exposed Region Detection -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Dataset -- 3.2 Network Architecture -- 3.3 Implementation -- 4 Experimental Results and Discussion -- 5 Conclusion -- References -- DeepHDR-GIF: Capturing Motion in High Dynamic Range Scenes -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 LDR Stack Generation -- 3.2 Dividing Images in LDR Set -- 3.3 HDR Stack Generation -- 3.4 Intermediate Frame Generation -- 4 Result -- 5 Limitations and Future Work -- 6 Conclusions -- References -- Camera Based Parking Slot Detection for Autonomous Parking -- 1 Introduction -- 2 Camera Setup and Calibration -- 3 Parking Slot Marks Detection -- 4 Occupancy Check -- 4.1 Motion Calculation -- 4.2 Homography Guided Reference Motion Estimation -- 4.3 Obstacle Detection Using Homography Guided Motion Segmentation (HGMS) -- 5 Results -- 6 Conclusions -- References. ; Hard-Mining Loss Based Convolutional Neural Network for Face Recognition -- 1 Introduction -- 2 Proposed Hard-Mining Loss -- 2.1 Cross-Entropy Loss -- 2.2 Hard-Mining Loss -- 2.3 Hard-Mining Cross-Entropy Loss -- 2.4 Hard-Mining Angular-Softmax Loss -- 2.5 Hard-Mining ArcFace Loss -- 3 Experimental Setup -- 3.1 CNN Architectures -- 3.2 Training Datasets -- 3.3 Testing Datasets -- 3.4 Input Data and Network Settings -- 4 Experimental Results and Observations -- 5 Conclusion -- References -- Domain Adaptive Egocentric Person Re-identification -- 1 Introduction -- 2 Related Work -- 2.1 Classic Person Re-ID -- 2.2 Egocentric Person Re-ID -- 2.3 Domain Adaptation in Image-to-Image Translation and Person Re-ID -- 3 Approach -- 3.1 Content Loss -- 3.2 Style Loss -- 3.3 Total Loss -- 3.4 Style Transfer Based Person Re-ID -- 4 Experiments and Results -- 4.1 Datasets -- 4.2 Evaluation Methodologies -- 4.3 Implementation Details -- 4.4 Results -- 5 Conclusion -- References -- Scene Text Recognition in the Wild with Motion Deblurring Using Deep Networks -- 1 Introduction -- 2 Related Work -- 3 The Proposed Methodology -- 3.1 Textbox Detection -- 3.2 Deblurring -- 3.3 Text Recognition -- 4 Experiments -- 4.1 Datasets -- 4.2 Training Details -- 4.3 Results -- 5 Conclusion and Future Work -- References -- Vision Based Autonomous Drone Navigation Through Enclosed Spaces -- 1 Introduction -- 2 Overview -- 2.1 Hardware Setup -- 2.2 Software Setup -- 3 Our Approach -- 3.1 Localisation -- 3.2 Entry Point Detection and Trajectory Generation -- 3.3 Autonomous Navigation -- 4 Experimentation and Results -- 4.1 Accuracy of Localisation Module -- 4.2 Accuracy for Entry Point Detection -- 5 Conclusion -- References -- Deep Learning-Based Smart Parking Management System and Business Model -- Abstract -- 1 Introduction -- 2 System Model -- 2.1 System Architecture. ; 3 Proposed System -- 3.1 OCR -- 3.2 Localization and Segmentation -- 3.3 Pre-processing and Feature Extraction -- 3.4 Recognition of Alphanumeric -- 3.5 TensorFlow Lite -- 3.6 Hardware -- 3.6.1 Raspberry Pi -- 3.6.2 Noir V2 Pi Camera -- 4 Implementation -- 4.1 Mobile Application -- 4.2 Free Space Identification -- 4.3 Authenticating User Vehicle -- 4.4 Notifying the User -- 5 Business Model -- 6 Experimental Results -- 7 Conclusion -- 8 Limitations -- References -- Design and Implementation of Motion Envelope for a Moving Object Using Kinect for Windows -- Abstract -- 1 Introduction -- 2 Background -- 2.1 Motion Envelope -- 2.2 Kinect Sensor -- 2.3 Feature Based Object Detection -- 2.4 Object Localisation Using Homography -- 3 Design and Implementation -- 3.1 Object Detection and Tracking -- 3.2 Post Processing -- 4 Results and Discussion -- 4.1 Determine Motion Envelope -- 4.2 Evaluation of Object Detection Algorithm -- 4.3 Evaluation of the Designed Algorithm -- 5 Conclusion and Future Work -- References -- Software Auto Trigger Recording for Super Slow Motion Videos Using Statistical Change Detection -- Abstract -- 1 Introduction -- 2 Method -- 2.1 Proposed Method -- 3 Experiments and Results -- 3.1 Qualitative Analysis -- 3.2 Quantitative Analysis -- 4 Conclusion -- References -- Using Class Activations to Investigate Semantic Segmentation -- 1 Introduction -- 1.1 Our Contribution -- 2 Background Formulation -- 2.1 Semantic Segmentation -- 2.2 Class Activation Maps -- 2.3 GradCAM -- 2.4 Guided GradCAM -- 3 Proposed Framework -- 3.1 Key Features -- 3.2 Segmentation Subnetwork -- 3.3 Masking Images -- 3.4 Classifier -- 4 Sanity Checks for Saliency Maps -- 5 Experiments -- 5.1 Segmentation Subnetwork -- 5.2 Masked Image Classifier -- 5.3 Overall Analysis -- 6 Results and Discussions -- 7 Conclusion -- 8 Future Work -- References. ; Few Shots Learning: Caricature to Image Recognition Using Improved Relation Network -- 1 Introduction and Related Work -- 2 Methodology -- 2.1 Problem Formulation -- 2.2 Proposed Model -- 2.3 Testing -- 2.4 Network Architecture -- 3 Experiments and Results -- 3.1 Dataset Description -- 3.2 Experimental Settings -- 3.3 Results and Discussion -- 4 Conclusion -- References -- Recognition of Adavus in Bharatanatyam Dance -- 1 Introduction -- 2 Related Work -- 2.1 Key Frame Extraction -- 2.2 Recognition of Key Postures -- 2.3 Recognition of Dance Sequence or Adavu -- 3 Data Set -- 3.1 Data Set Extension - Video Sub-Sequences -- 4 System Architecture -- 5 Feature Extraction and Recognizers -- 5.1 Feature Extraction -- 5.2 Recognition of KP Using SVM -- 5.3 Adavu Recognition using SVM and Edit Distance -- 6 Conclusions -- References -- Digital Borders: Design of an Animal Intrusion Detection System Based on Deep Learning -- 1 Introduction -- 2 Related Work -- 2.1 Animal Intrusion Detection Systems -- 2.2 Animal Re-identification -- 3 Proposed Methodology -- 3.1 Animal Detection -- 3.2 Animal Re-identification Network -- 3.3 System Design -- 4 Datasets -- 5 Results and Discussion -- 6 Future Work and Conclusion -- References -- Automatic On-Road Object Detection in LiDAR-Point Cloud Data Using Modified VoxelNet Architecture -- 1 Introduction -- 2 Literature Review -- 2.1 Bird's Eye View Based Method -- 2.2 Front-View and Image-Based Method -- 2.3 Fusion-Based Methods -- 2.4 3D-Based Methods -- 3 Background for the Proposed Architecture -- 3.1 Feature Encoder in Point Cloud -- 3.2 The Backbone Network -- 3.3 Object Detection in Point Cloud -- 4 Proposed Models -- 4.1 Method-1 -- 4.2 Method-2 -- 4.3 Method-3 -- 5 Experimental Results -- 6 Conclusion -- References -- On the Performance of Convolutional Neural Networks Under High and Low Frequency Information. ; 1 Introduction -- 2 High and Low Pass Image Dataset Preparation -- 2.1 New Test Sets Generation Using High and Low Pass Filtering -- 2.2 Training Set Augmentation Using Stochastic Filtering -- 3 Network Architecture and Training Settings -- 3.1 Residual Network -- 3.2 Training Settings -- 4 Experimental Results and Discussion -- 4.1 Experiment-1 -- 4.2 Experiment-2 -- 4.3 Experiment-3 -- 4.4 Experiment-4 -- 5 Conclusion -- References -- A Lightweight Multi-label Image Classification Model Based on Inception Module -- 1 Introduction -- 2 Related Work -- 3 The Proposed Method -- 4 Results and Discussion -- 4.1 Datasets Used in the Experiment -- 4.2 Performance Metrics -- 4.3 Experimental Results and Comparative Study -- 5 Conclusion and Future Work -- References -- Computer Vision based Animal Collision Avoidance Framework for Autonomous Vehicles -- 1 Introduction -- 2 Methodology -- 2.1 Animal Detection -- 2.2 Lane Detection -- 2.3 Animal Direction and Vicinity Tracking -- 2.4 Overall Pipeline -- 3 Experimental Results -- 3.1 Datasets -- 3.2 Results -- 4 Conclusions and Future Scope -- References -- L2PF - Learning to Prune Faster -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Problem Formulation -- 3.2 Environment -- 3.3 Distinct Action Space for Pruning and Epoch-Learning -- 3.4 Multi-objective Reward Function -- 3.5 Agent Design -- 4 Experimental Results -- 4.1 Design Space Exploration -- 4.2 Class Activation Maps -- 4.3 Comparison with the State-of-the-Art -- 5 Conclusion -- References -- Efficient Ensemble Sparse Convolutional Neural Networks with Dynamic Batch Size -- 1 Introduction -- 2 Related Work -- 2.1 Convolutional Accelerator -- 2.2 Activation Function -- 2.3 Batch Size -- 3 Proposed Method -- 3.1 Stacking -- 3.2 Winograd-ReLU CNN -- 3.3 Strategies for Better Convergence -- 3.4 Dynamic Batch Size -- 4 Experiments. ; 4.1 FASHION MNIST (AlexNet).
• Titolo autorizzato: Computer vision and image processing
• ISBN: 981-16-1103-3
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 996464391803316
• Lo trovi qui: Univ. di Salerno
• Serie: Communications in computer and information science ; ; 1378.