14 papers
Shared Geometry As A Rosetta Stone: Cross-Modal Alignment Without Paired Data
Dominik Schnaus, Thomas Dagès, Daniel Cremers +2
Multimodal representations enable zero-shot classification and retrieval, but aligning independently trained models usually requires large amounts of paired data. Yet, the Platonic…
Permutation-Equivariant Flow Matching for Alignment-Free Neural Weight Generation
Arkadi Piven, Yam Eitan, Guy Bar-Shalom +5
A trained neural network can be represented by a parameter vector in high dimensions. Learning distributions over these vectors enables the generation of new models across various…
Initialization-Free Bundle Adjustment Revisited: A Controlled Experimental Study
Simon Weber, Mateo de Mayo, Je Hyeong Hong +3
Initialization-free bundle adjustment (InitFree BA) aims to recover camera poses and scene structure directly from image observations, avoiding the geometric initialization stages…
TunerDiT: Training-free Progressive Steering of Diffusion Transformer for Multi-Event Video Generation
Ruotong Liao, Guowen Huang, Qing Cheng +6
Text-to-video (T2V) generation faces challenging questions when generating videos with long horizons containing multiple events. Inspired by the intrinsics of the diffusion process…
Functional Attention: From Pairwise Affinities to Functional Correspondences
Jiefang Xiao, Maolin Gao, Simon Weber +2
Learning mappings between infinite-dimensional function spaces, or operator learning, is essential for many machine learning applications. Although transformer-based operators are…
Graph Neural Networks Are Not Continuous Across Graph Resolutions
Christian Koke, Yuesong Shen, Abhishek Saroha +4
We show that contrary to conventional wisdom in the community, graph neural networks (GNNs) are not continuous with respect to all natural modes of graph convergence. As a result,…
Generative Shape Reconstruction with Geometry-Guided Langevin Dynamics
Linus Härenstam-Nielsen, Dmitrii Pozdeev, Thomas Dagès +2
Reconstructing complete 3D shapes from incomplete or noisy observations is a fundamentally ill-posed problem that requires balancing measurement consistency with shape plausibility…
Harnessing Data Asymmetry: Manifold Learning in the Finsler World
Thomas Dagès, Simon Weber, Daniel Cremers +1
Manifold learning is a fundamental task at the core of data analysis and visualisation. It aims to capture the simple underlying structure of complex high-dimensional data by prese…
Learning Eigenstructures of Unstructured Data Manifolds
Roy Velich, Arkadi Piven, David Bensaïd +3
We introduce a novel framework that directly learns a spectral basis for shape and manifold analysis from unstructured data, eliminating the need for traditional operator selection…
FlowFeat: Pixel-Dense Embedding of Motion Profiles
Nikita Araslanov, Anna Sonnweber, Daniel Cremers
Dense and versatile image representations underpin the success of virtually all computer vision applications. However, state-of-the-art networks, such as transformers, produce low-…
Finsler Multi-Dimensional Scaling: Manifold Learning for Asymmetric Dimensionality Reduction and Embedding
Thomas Dagès, Simon Weber, Ya-Wei Eileen Lin +5
Dimensionality reduction is a fundamental task that aims to simplify complex data by reducing its feature dimensionality while preserving essential patterns, with core applications…
Power Variable Projection for Initialization-Free Large-Scale Bundle Adjustment
Simon Weber, Je Hyeong Hong, Daniel Cremers
Most Bundle Adjustment (BA) solvers like the Levenberg-Marquardt algorithm require a good initialization. Instead, initialization-free BA remains a largely uncharted territory. The…
Flattening the Parent Bias: Hierarchical Semantic Segmentation in the Poincaré Ball
Simon Weber, Barış Zöngür, Nikita Araslanov +1
Hierarchy is a natural representation of semantic taxonomies, including the ones routinely used in image segmentation. Indeed, recent work on semantic segmentation reports improved…
Finsler-Laplace-Beltrami Operators with Application to Shape Analysis
Simon Weber, Thomas Dagès, Maolin Gao +1
The Laplace-Beltrami operator (LBO) emerges from studying manifolds equipped with a Riemannian metric. It is often called the Swiss army knife of geometry processing as it allows t…