Abstract:Leo Tolstoy opened his monumental novel Anna Karenina with the now famous words: Happy families are all alike; every unhappy family is unhappy in its own way A similar notion also applies to mathematical spaces: Every flat space is alike; every unflat space is unflat in its own way. However, rather than being a source of unhappiness, we will show that the diversity of non-flat spaces provides a rich area of study. The genesis of the so-called big data era and the proliferation of social and scientific databases of increasing size has led to a need for algorithms that can efficiently process, analyze and, even generate high dimensional data. However, the curse of dimensionality leads to the fact that many classical approaches do not scale well with respect to the size of these problems. One technique to avoid some of these ill-effects is to exploit the geometric structure of coherent data. In this thesis, we will explore geometric methods for shape processing and data analysis. More specifically, we will study techniques for representing manifolds and signals supported on them through a variety of mathematical tools including, but not limited to, computational differential geometry, variational PDE modeling, and deep learning. First, we will explore non-isometric shape matching through variational modeling. Next, we will use ideas from parallel transport on manifolds to generalize convolution and convolutional neural networks to deformable manifolds. Finally, we conclude by proposing a novel auto-regressive model for capturing the intrinsic geometry and topology of data. Throughout this work, we will use the idea of computing correspondences as a though-line to both motivate our work and analyze our results.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is the "curse of dimensionality" in high - dimensional data processing and how to effectively use the geometric structure of data to design new algorithms. Specifically, the author explores the application of geometric methods in shape processing and data analysis, especially when dealing with data in non - Euclidean spaces. The main contributions of the paper can be summarized as follows:
1. **Non - isometric shape matching**: Explore the matching problem between non - isometric shapes through variational modeling methods. This part mainly focuses on how to find the correspondence between two different shapes, even if these shapes have undergone large deformations.
2. **Parallel transport convolution**: Utilize the concept of parallel transport on manifolds to generalize convolution and convolutional neural networks to deformable manifolds. This enables the convolution operation to be effectively carried out in non - Euclidean spaces, thereby better handling complex data structures.
3. **Chart - based auto - encoder**: Propose a new autoregressive model for capturing the intrinsic geometry and topological structure of data. This method can not only effectively represent data but also maintain the geometric properties of data when generating new data.
### Specific problems and their solutions
#### 1. Non - isometric shape matching
- **Problem**: How to find corresponding points between two non - isometrically deformed shapes?
- **Solution**: Use the Variational PDE Model and regularization techniques, combined with area - constraint conditions, to construct an optimization problem to solve the correspondence between non - isometric shapes. Specifically, find the best correspondence between shapes by solving the eigenvalue problem of the Laplace - Beltrami Operator.
#### 2. Parallel transport convolution
- **Problem**: How to define and implement the convolution operation in non - Euclidean spaces?
- **Solution**: Utilize the concept of Parallel Transportation on manifolds to define Parallel Transport Convolution (PTC). Through numerical discretization methods, apply PTC to actual manifold data, realizing the effective generalization of convolutional neural networks in non - Euclidean spaces.
#### 3. Chart - based auto - encoder
- **Problem**: How to capture the intrinsic geometry and topological structure of data and generate new data?
- **Solution**: Propose a Chart Auto - Encoders (CAE) based on charts, represent the low - dimensional structure of data through multiple local charts. Each chart is responsible for representing a part of the data, and through the transformation functions between charts, the overall representation of data is achieved. In addition, by introducing a regularization term, the geometric and topological consistency of the generated new data is ensured.
### Main contributions of the paper
- **Theoretical contributions**: Propose new mathematical tools and methods, such as the Variational PDE Model, Parallel Transport Convolution, and Chart - based auto - encoder, providing a new theoretical basis for processing high - dimensional data.
- **Practical contributions**: Verify the effectiveness of the proposed methods through a large number of experiments, demonstrating the potential of these methods in practical applications such as shape matching, image processing, and data generation.
In general, this paper aims to solve the challenges in high - dimensional data processing through geometric methods, especially how to effectively use the intrinsic geometric structure of data to design new algorithms. These methods are not only innovative in theory but also show strong performance in practical applications.