Intelligent Scene Modeling and Human-Computer Interaction

Preface
Contents
Contributors
Part I Intelligent Scene Modeling (ISM)
1 Introduction
2 Object Detection: State of the Art and Beyond
	2.1 Introduction
	2.2 Preliminary
		2.2.1 Definition
		2.2.2 General Framework of DNN-Based Object Detection
		2.2.3 Datasets
		2.2.4 Evaluation Metrics
		2.2.5 Open-Source Libraries for Object Detection
	2.3 Exemplar Baselines
		2.3.1 Two-Stage Detectors
		2.3.2 One-Stage Detectors
	2.4 Cutting-Edge Detectors and Trends
		2.4.1 Feature Fusion
		2.4.2 Anchor-Free Detectors
	2.5 Conclusion
	References
3 NBNN-Based Discriminative 3D Action and Gesture Recognition
	3.1 Introduction
		3.1.1 Related Work
	3.2 Proposed Method
		3.2.1 3D Action Representation
		3.2.2 Classification
	3.3 Experiments
		3.3.1 Implementation Details
		3.3.2 Results and Analysis
	3.4 Conclusions
	3.5 Future Work
	References
4 Random Forests with Optimized Leaves for Hough-Voting
	4.1 Introduction
	4.2 Literature Review
	4.3 The Proposed Method
		4.3.1 Leaf Selection and Weight Optimization
		4.3.2 Details of Forest Learning and Prediction
	4.4 Experimental Results
		4.4.1 Hand Pose Estimation in Depth Images
		4.4.2 Head Pose Estimation in Depth Images
		4.4.3 Hand Gesture Recognition in Depth Images
	4.5 Conclusion
	References
5 Modeling Human Perception of 3D Scenes
	5.1 Introduction
	5.2 Visual Attention
		5.2.1 Artificial Intelligence for Visual Attention
		5.2.2 Future Directions on Visual Attention
	5.3 Visual Quality Perception of 3D Shapes
		5.3.1 Artificial Intelligence for Visual Quality Assessment
		5.3.2 Future Directions on 3D VQA
	5.4 Visual Material Perception and Recognition
		5.4.1 Materials in Computer Graphics
		5.4.2 Visual Material Perception
		5.4.3 Artificial Intelligence for Material Recognition
		5.4.4 Existing Material Datasets
		5.4.5 Future Directions on Visual Material Perception
	References
6 Model Reconstruction of Real-World 3D Objects: An Application with Microsoft HoloLens
	6.1 Introduction
	6.2 Microsoft HoloLens
		6.2.1 Application Programming Interfaces (API)
	6.3 Model Reconstruction Pipeline
		6.3.1 Pixel Ray-Casting for Texture Mapping
		6.3.2 Mesh and Texture Rendering
		6.3.3 Texture Pixel Dilation
	6.4 Results and Discussions
		6.4.1 3D Object Model Reconstruction
		6.4.2 Computation
	6.5 Conclusions and Future Work
	References
7 Semantic Modeling and Rendering
	7.1 Introduction
	7.2 Semantic Modeling
		7.2.1 Representation of Shape Semantics
		7.2.2 Capture and Modeling Shape Semantics
		7.2.3 Applications of Shape Semantics
		7.2.4 Scheme for Generating 3D Models
	7.3 Examples: Rendering Complex Urban Model
		7.3.1 Semantic Data Structure
		7.3.2 Database for Large-Scale Semantic Urban Rendering
		7.3.3 Integration of Application-Specific Semantics for 3D Smart City Services
	7.4 Conclusions
	References
8 Content-Aware Semantic Indoor Scene Modeling from a Single Image
	8.1 Introduction and Related Work
		8.1.1 Indoor Content Parsing
		8.1.2 Support Inference
		8.1.3 Scene Modeling
	8.2 Overview
	8.3 Scene Parsing
		8.3.1 Instance Segmentation
		8.3.2 Depth Estimation
		8.3.3 Layout Estimation
	8.4 Support Inference
		8.4.1 Support Inference from Priors
		8.4.2 Support Inference with Relation Network
	8.5 Scene Modeling
		8.5.1 Model Matching Formulation
		8.5.2 Global-to-Local Model Search and Refinement
	8.6 Experiment and Discussions
		8.6.1 Performance Analysis
		8.6.2 Discussions
	8.7 Conclusions
	References
9 Interactive Labeling for Generation of CityGML Building Models from Meshes
	9.1 Introduction
	9.2 Related Work
	9.3 Overview of the Proposed Approach
		9.3.1 Wavefront OBJ Models
		9.3.2 CityGML Models
		9.3.3 Outline of the Algorithm
	9.4 Geometric Processing for Interactive Labeling
		9.4.1 Outer Hull Extraction
		9.4.2 Face Adjacency Graph
		9.4.3 Criteria for Semantic Grouping
		9.4.4 Graph-Cut-Based Selection
	9.5 Experiments
	9.6 Conclusion
	References
10 TooltY: An Approach for the Combination of Motion Capture and 3D Reconstruction to Present Tool Usage in 3D Environments
	10.1 Introduction
	10.2 Related Work
		10.2.1 Key Technologies
		10.2.2 Motion Visualization
		10.2.3 Authoring Tools for Experiences and Tool Usage Demonstration
		10.2.4 Embodied Agents—Virtual Humans
	10.3 The Proposed Pipeline: From Motion Capture to Motion Visualization
	10.4 Results and Contribution
	10.5 Future Work
	References
11 Generating 3D Facial Expressions with Recurrent Neural Networks
	11.1 Introduction
	11.2 Facial Shape Modeling
		11.2.1 Facial Shape Space
		11.2.2 Other Shape Representations
		11.2.3 Modern Facial Modelers Using DNN
	11.3 Facial Animation Modeling
	11.4 Learning-Based Generation of 3D Facial Expression Sequences
		11.4.1 RNN on Time Series Data
		11.4.2 Learning to Generate 3D Facial Animations
	11.5 Discussion
	References
Part II Human Computer Interaction (HCI)
12 Facilitating Decision-Making with Multimodal Interfaces in Collaborative Analytical Sessions
	12.1 Introduction
	12.2 Related Work
		12.2.1 Collaborative Visual Analytics
		12.2.2 Natural Language Interfaces
		12.2.3 Natural Language Interfaces for Visual Analytics
	12.3 System Analysis
		12.3.1 Analytical Agent
		12.3.2 Visual Analytics System
	12.4 Data Overview
	12.5 System Description
		12.5.1 Analytical Agent
		12.5.2 Visual Analytics System
	12.6 Qualitative Evaluation
		12.6.1 Evaluation with Novice Users (C1)
		12.6.2 Evaluation with Domain Experts (C2)
	12.7 Discussion
	12.8 Conclusion
	References
13 Human—Technology Interaction: The State-of-the-Art and the Lack of Naturalism
	13.1 Introduction
	13.2 Human Perception During HCI
		13.2.1 The Role of Human Likeness
		13.2.2 The Role of Embodiment and Presence
		13.2.3 Other Features that Can Influence Human’s Perception During HCI
	13.3 Human—Robot Interaction (HRI)
		13.3.1 Social Robots and Their Features
	13.4 Humans and Virtual Avatars
		13.4.1 Conceptualization and Perception of Avatars
	13.5 Conclusions
	References
14 Survey of Speechless Interaction Techniques in Social Robotics
	14.1 Introduction
	14.2 Related Work
	14.3 Gaze Behavior
	14.4 Affective System—Emotion Engine
	14.5 Recognize and React to Actions/Gestures
	14.6 Natural Language Communication with Robots Using Reading
	14.7 Internet and the HRI
	14.8 Haptics in Social Robotics
	14.9 Conclusion
	References
15 Exploring Potential and Acceptance of Socially Intelligent Robot
	15.1 Introduction
	15.2 State of the Art
		15.2.1 Robot as Interviewer
		15.2.2 Robots as Teacher
		15.2.3 Robot as Customer Guide
		15.2.4 Robot as Companion
	15.3 Research Questions
	15.4 Materials and Methods
		15.4.1 Nadine Social Robot
		15.4.2 Participants
		15.4.3 Experimental Design
		15.4.4 NARS
		15.4.5 Psychometric
		15.4.6 Body Movements
		15.4.7 EEG Recordings and Analysis
		15.4.8 Statistics
	15.5 Results
		15.5.1 NARS Data
		15.5.2 Psychometric Data
		15.5.3 Kinect Data
		15.5.4 EEG Data
	15.6 Discussion and Future Work
	15.7 Conclusion
	References