Sparse coding based image restoration and recognition algorithms and analysis

Contents

List of Figures

List of Tables

1 Introduction

1.1 Background
- 1.1.1 Dictionary learning for image restoration and recognition
- 1.1.2 Dictionary learning algorithms
- 1.1.3 Proximal methods
1.2 Motivations and contributions of the dissertation
- 1.2.1 Data-driven tight frame construction
- 1.2.2 Redundant dictionary learning
- 1.2.3 Incoherent dictionary learning
- 1.2.4 L1 visual tracker
1.3 Notation

2 Data-driven tight frame construction for image restoration

2.1 Introduction
2.2 Brief review on data-driven tight frame construction and related works
- 2.2.1 Tight frames and data-driven tight frames
- 2.2.2 Data-driven tight frame construction scheme
- 2.2.3 Related works
2.3 Sub-sequence convergence property of Algorithm 1
2.4 A modified algorithm for (2.7) with sequence convergence
- 2.4.1 Convergence analysis of Algorithm 2
2.5 Experiments on image denoising
2.6 Extensions
- 2.6.1 Problem formulation
- 2.6.2 Numerical method
- 2.6.3 Complexity analysis of Algorithm 3
- 2.6.4 Applications in image restoration
- 2.6.5 Experiments
- 2.6.6 Discussion and conclusion

3 Redundant dictionary learning for image restoration and recognition

3.1 Introduction
- 3.1.1 Motivation
- 3.1.2 Main contributions
3.2 Related work
- 3.2.1 0 norm based methods
- 3.2.2 Convex relaxation methods
- 3.2.3 Non-convex relaxation methods
3.3 Algorithm and convergence analysis
- 3.3.2 Alternating proximal method
3.4 Global convergence of Algorithm 6
3.5 Experiments
- 3.5.1 Image denoising
- 3.5.2 Face recognition
3.6 Summary

4 Incoherent dictionary learning for image recognition

4.1 Introduction
- 4.1.1 Motivation and main contributions
- 4.1.2 Related work
4.2 Incoherent dictionary learning algorithm
- 4.2.2 A hybrid alternating proximal algorithm
4.3 Convergence analysis of Algorithm 7
4.4 Experiments
- 4.4.1 Experimental setting
- 4.4.2 Experimental results
4.5 Summary and conclusions

5 Sparse coding based visual tracking

5.1 Introduction
5.2 Related work
5.3 Introduction to L1 Tracker
5.4 Real time L1 Tracker
- 5.4.1 A modified 1 norm related minimization model
- 5.4.2 Fast numerical method for solving (5.9)
5.5 Experiments
- 5.5.1 Comparison with the existing L1 Tracker
- 5.5.2 Qualitative comparison with other methods
- 5.5.3 Quantitative comparison with other methods
5.6 Conclusion

Bibliography