AG 1, AG 2, AG 3, INET, AG 4, AG 5, D6, SWS, RG1, MMCI   Info

AG Audience

English

30 Minutes

In this talk, I will introduce NeuralSSD, a solver based on the neural Galerkin method designed to reconstruct high-quality and accurate 3D implicit surfaces from widely available point cloud data. Implicit methods are particularly preferred due to their ability to represent complex shapes accurately and their robustness in handling topological optimizations. During this project, we identified the effectiveness of sparse voxel hierarchies in representing the sparsity of 3D data while preserving multi-scale details critical for geometric fidelity. Building on this insight, I am currently exploring new representations for dynamic scenes. we focus on representing motion fields and 3D Gaussian in dynamic 4D settings to address the challenges of reconstructing dynamic scenes from image and video data.

Ina Geisler

+49 681 9325 1802

--email hidden

Zoom

passcode not visible

logged in users only