Intro
Preface
Organization
Contents
Part II
Intelligent Computing in Computer Vision
BIDGAN: Blind Image Deblurring with Improved CycleGAN and Frequency Filtering
Abstract
1 Introduction
2 Blind Image Deblurring Based on CycleGAN
2.1 Generative Adversarial Network
2.2 CycleGAN Based on Wasserstein Loss
2.3 Generator Network
2.4 Discriminator Network
2.5 Switchable Normalization for Motion Blur Removing
3 Experiments
3.1 GOPRO Dataset
3.2 Model Training
3.3 Experiment Results
4 Image Frequency Domain Analysis
5 Conclusion
References

Emotional Interaction Computing of Actors in the Mass Incidents
Abstract
1 Theoretical Basis of Group Psychology Research
1.1 Emotional Infection Mechanism
1.2 Emotional Characteristics of Participants in Mass Incidents
2 Computational Model of Emotional Interaction Among Participants in Mass Events
3 Model Simulation and Empirical Analysis
3.1 Review of the Death of a Woman in Jingwen Mall
3.2 Model Simulation
3.2.1 The Real Crowd Simulation
3.2.2 Network Crowd Simulation
3.2.3 Emotional Spiral in Emotional Infection
4 Conclusion
References

Multi Spatial Convolution Block for Lane Lines Semantic Segmentation
Abstract
1 Introduction
2 Related Work
3 Method
4 Experiment
5 Conclusion
Acknowledgments
References
VISFF: An Approach for Video Summarization Based on Feature Fusion
Abstract
1 Introduction
2 Related Works
3 Video Summarization Based on Feature Fusion
3.1 Video Frame Preprocessing
3.2 Feature Extraction
3.3 Feature Fusion
3.4 DBSCAN Clustering
3.5 Keyframe Extraction
4 Experimental Results
4.1 Datasets
4.2 Evaluation Method
4.3 Evaluation on OVP Dataset

4.4 Evaluation on YouTube Dataset
5 Conclusion
Acknowledgment
References
Understanding Safety Based on Urban Perception
Abstract
1 Introduction
2 Related Works
2.1 Urban Perception
2.2 Model Interpretation.
3 Methodology
3.1 Dataset
3.2 Experiments
3.3 Model Explanation
4 Discussions
5 Conclusions
References
Recognition of Multiple Panamanian Watermelon Varieties Based on Feature Extraction Analysis
Abstract
1 Introduction
2 Proposed Method
2.1 Image Preprocessing
2.2 Candidate Detection
2.3 Classification
2.4 Classification

3 Experimental Results
3.1 Edge Detection Evaluation
3.2 Ellipsoid Detection
3.3 Fitting Estimation
4 Conclusions and Future Work
Acknowledgment
References
STDA-inf: Style Transfer for Data Augmentation Through In-data Training and Fusion Inference
Abstract
1 Introduction
2 Related Work
2.1 Style Transfer
2.2 Data Augmentation
3 Proposed Approach
3.1 In-data Training
3.2 Fusion Inference
4 Experiments
4.1 Dataset
4.2 Implementation Details
4.3 Experiment Results
5 Conclusion
6 Appendix
Acknowledgment
References