Course Identification

Linear algebra, calculus. 

This is a student seminar that provides introduction to geometry processing and the emerging field of geometric deep learning.

The Course will consists of two parts:

1. introduction to discrete differential geometry (given by Yaron).

2. seminar lectures (given by participating students) in:

1. Geometry processing.

2. Geometric deep learning.

Each student is expected to attend all classes and prepare and present one lecture on one of the topics below (see seminar part). Furthermore, weekly assignment will be to read and summarize 1-3 papers in a short paragraph.

Introductory part (delivered by Yaron 2-3 meetings):

1. Introduction to geometry processing.

   1. Image processing - geometry processing.

   2. Surfaces, meshes.

   3. Piecewise linear functions, gradients, dirichlet energy, discrete harmonic.

   4. Curvature, Euler characteristic, Gauss-Bonnet, Vector fields, deformations.

   5. Applications:

      1. Processing: smoothing, morphing, reconstruction, deformations, parameterizations, mappings.

      2. Analysis: shape matching and distance, structure analysis (symmetry, inverse engineering), shape space.

Seminar part (delivered by participating students):

2. Geometry processing.

   1. Computational topology.

   2. Discrete differential geometry.

   3. Tutte embedding.

   4. Quadratic matching.

   5. Spectral shape analysis.

   6. Cross fields.

   7. Deformations.

3. Geometric deep learning.

   1. Rendering based methods.

   2. Spectral deep learning.

   3. Patch based methods.

   4. Parameterization based methods.

   5. Point nets.

   6. Shape generation.

Upon successful completion of this course students will be able to:

• Get familiar with current research in geometric deep learning.