( events)   Timezone: »  
Fri Dec 11 05:45 AM -- 02:00 PM (PST)
Differential Geometry meets Deep Learning (DiffGeo4DL)
Joey Bose · Emile Mathieu · Charline Le Lan · Ines Chami · Fred Sala · Christopher De Sa · Maximillian Nickel · Chris Ré · Will Hamilton

Workshop Home Page

Recent years have seen a surge in research at the intersection of differential geometry and deep learning, including techniques for stochastic optimization on curved spaces (e.g., hyperbolic or spherical manifolds), learning embeddings for non-Euclidean data, and generative modeling on Riemannian manifolds. Insights from differential geometry have led to new state of the art approaches to modeling complex real world data, such as graphs with hierarchical structure, 3D medical data, and meshes.
Thus, it is of critical importance to understand, from a geometric lens, the natural invariances, equivariances, and symmetries that reside within data.

In order to support the burgeoning interest of differential geometry in deep learning, the primary goal for this workshop is to facilitate community building and to work towards the identification of key challenges in comparison with regular deep learning, along with techniques to overcome these challenges. With many new researchers beginning projects in this area, we hope to bring them together to consolidate this fast-growing area into a healthy and vibrant subfield. In particular, we aim to strongly promote novel and exciting applications of differential geometry for deep learning with an emphasis on bridging theory to practice which is reflected in our choices of invited speakers, which include both machine learning practitioners and researchers who are primarily geometers.

gather.town (Social)
Opening Remarks (Presentation)
Invited Talk 1: Geometric deep learning for 3D human body synthesis (Talk by Michael Bronstein)
Invited Talk 2: Gauge Theory in Geometric Deep Learning (Talk by Taco Cohen)
Contributed Talk 1: Learning Hyperbolic Representations for Unsupervised 3D Segmentation (Contributed Talk)
Contributed Talk 2: Witness Autoencoder: Shaping the Latent Space with Witness Complexes (Contributed Talk 2)
Contributed Talk 3: A Riemannian gradient flow perspective on learning deep linear neural networks (Contributed Talk 3)
Contributed Talk 4: Directional Graph Networks (Contributed Talk 4)
Contributed Talk 5: A New Neural Network Architecture Invariant to the Action of Symmetry Subgroups (Contributed Talk 5)
Virtual Coffee Break on Gather.Town (Break)
Invited Talk 3: Reparametrization invariance in representation learning (Talk by Søren Hauberg)
Grassmann Iterative Linear Discriminant Analysis with Proxy Matrix Optimization (Poster)
Graph of Thrones : Adversarial Perturbations dismantle Aristocracy in Graphs (Poster)
GENNI: Visualising the Geometry of Equivalences for Neural Network Identifiability (Poster)
Tree Covers: An Alternative to Metric Embeddings (Poster)
Quaternion Graph Neural Networks (Poster)
Universal Approximation Property of Neural Ordinary Differential Equations (Poster)
Hermitian Symmetric Spaces for Graph Embeddings (Poster)
Isometric Gaussian Process Latent Variable Model (Poster)
A Metric for Linear Symmetry-Based Disentanglement (Poster)
Deep Networks and the Multiple Manifold Problem (Poster)
Poster Session 1 on Gather.Town (Poster Session)
Panel Discussion (Panel)
Virtual Coffee Break on Gather.Town (Break)
Focused Breakout Session Companion Notebook: Poincare Embeddings (Demonstration)
Focused Breakout Session (Demonstration)
Focused Breakout Session Companion Notebook: Wrapped Normal Distribution (Demonstration)
Invited Talk 4: An introduction to the Calderon and Steklov inverse problems on Riemannian manifolds with boundary (Talk by Niky Kamran)
Sparsifying networks by traversing Geodesics (Poster)
Deep Riemannian Manifold Learning (Poster)
The Intrinsic Dimension of Images and Its Impact on Learning (Poster)
Unsupervised Orientation Learning Using Autoencoders (Poster)
Leveraging Smooth Manifolds for Lexical Semantic Change Detection across Corpora (Poster)
Poster Session 2 on Gather.Town (Poster Session)
QuatRE: Relation-Aware Quaternions for Knowledge Graph Embeddings (Poster)
Affinity guided Geometric Semi-Supervised Metric Learning (Poster)
Convex Optimization for Blind Source Separation on a Statistical Manifold (Poster)
Extendable and invertible manifold learning with geometry regularized autoencoders (Poster)
Towards Geometric Understanding of Low-Rank Approximation (Poster)
Invited Talk 5: Disentangling Orientation and Camera Parameters from Cryo-Electron Microscopy Images Using Differential Geometry and Variational Autoencoders (Talk by Nina Miolane)
Invited Talk 6: Learning a robust classifier in hyperbolic space (Talk by Melanie Weber)