Computer Vision -- ECCV 2020 : 16th European Conference, Glasgow, UK, August 23-28, 2020, Proceedings. Part VIII / Andrea Vedaldi, Horst Bischof, Thomas Brox, Jan-Michael Frahm (eds.).

Intro
Foreword
Preface
Organization
Contents
Part VIII
Weakly-Supervised Crowd Counting Learns from Sorting Rather Than Locations
1 Introduction
2 Related Work
2.1 Density Estimation Based Methods
2.2 Methods Dealing with the Lack of Labelled Data
2.3 Learning from Sorting
3 Method
3.1 Regression Network
3.2 Sorting Network
3.3 Training Method
4 Experiments
4.1 Implementation Details
4.2 Evaluation Metrics
4.3 Evaluation and Comparison
4.4 Ablation Study
5 Conclusions
References

Unsupervised Domain Attention Adaptation Network for Caricature Attribute Recognition
1 Introduction
2 Related Work
2.1 Face Attribute Recognition Methods
2.2 Unsupervised Domain Adaptation Methods in Classification
2.3 Image-to-Image Translation
3 The WebCariA Dataset
4 Method
4.1 The Unsupervised Domain Adaptation Framework
4.2 Inter-Domain Consistency Learning
4.3 Intra-domain Consistency Learning
4.4 Multi-task Attribute Recognition
4.5 Implementation Details
5 Experimental Results
5.1 Ablation Study
5.2 Comparison with the State-of-the-Art Methods

5.3 Visualization of the Attribute-Wise Attention Maps
6 Conclusions
References
Many-Shot from Low-Shot: Learning to Annotate Using Mixed Supervision for Object Detection
1 Introduction
2 Related Work
3 Method
3.1 Online Annotation Module
3.2 Fully Supervised Branch
4 Results
4.1 Datasets and Implementation Details
4.2 Comparisons with State-of-the-Art
4.3 Ablation Studies
5 Conclusion
References
Curriculum DeepSDF
1 Introduction
2 Related Work
3 Proposed Approach
3.1 Review of DeepSDF ch4park2019deepsdf
3.2 Curriculum DeepSDF

3.3 Implementation Details
4 Experiments
4.1 Shape Reconstruction
4.2 Missing Part Recovery
5 Conclusion
References
Meshing Point Clouds with Predicted Intrinsic-Extrinsic Ratio Guidance
1 Introduction
2 Related Work
2.1 Traditional Mesh Reconstruction
2.2 Learning-Based Mesh Generation
3 Method
3.1 A Motivating Remeshing Algorithm
3.2 From Remeshing to Reconstruction
4 Experiments
4.1 Data Generation and Network Training
4.2 Comparison with Existing Methods
4.3 Ablation Studies
5 Conclusions
References

Improved Adversarial Training via Learned Optimizer
1 Introduction
2 Related Work
2.1 Adversarial Attack and Defense
2.2 Learning to Learn
3 Preliminaries
3.1 Notations
3.2 Adversarial Training
3.3 Effects of Adaptive Step Sizes
4 Proposed Algorithm
4.1 Learning to Learn for Adversarial Training
4.2 Advantages over Other L2L-Based Methods
5 Experimental Results
5.1 Experimental Settings
5.2 Performance on White-Box Attacks
5.3 Analysis
5.4 Performance on Black-Box Transfer Attacks
5.5 Loss Landscape Exploration
6 Conclusion
References