Hosseini, Babak: Interpretable analysis of motion data. 2021
Inhalt
- Contents
- Introduction
- Foundations
- Motion Data Representation
- Formulating Motion Analysis Problems
- Measuring Motions Similarity by DTW
- Benchmark Motion Datasets
- Motion Data Analysis Literature
- Metric Learning for Motion Analysis
- State of The Art
- Distance-based Metric Learning
- Feasibility based Large Margin Nearest Neighbors
- Metric regularization
- Experiments
- Conclusion
- Sparse coding for Interpretable Embedding of Motion Data
- State of The Art
- Non-negative Kernel Sparse Coding
- Confidence based kernel Sparse Coding
- Motion Clustering using Non-negative Kernel Sparse Coding
- Experiments
- Conclusion
- Multiple Kernel Learning for Motion Analysis
- State of The Art
- Large-Margin Multiple Kernel Learning for Discriminative Feature Selection
- Interpretable Multiple-Kernel Prototype Learning
- Multiple-Kernel Dictionary Structure
- Experiments
- Conclusion
- Interpretable Motion Analysis with Convolutional Neural Network
- Conclusions and Outlook
- Publications in the Context of this Thesis
- References
- Appendix
- Proof of Theorem 3.1
- Proof of Lemma 3.1
- Additional Figures for Section 3.5
- Proof of Proposition 4.1
- The K-NNLS Algorithm
- Proof of Proposition 4.2
- Proof of Proposition 4.3
- Proof of Proposition 4.4
- Proof of Proposition 4.5
- Proof of Theorem 4.1
- Proof of Proposition 5.1
- Proof of Proposition 5.2
- Proof of Theorem 5.1
- Complete Architecture of DACNN from Section 6.3