Peter Kontschieder

29

Papers

143

Total Citations

Papers (29)

Revising Densification in Gaussian Splatting

Multi-Level Neural Scene Graphs for Dynamic Urban Environments

ConsistDreamer: 3D-Consistent 2D Diffusion for High-Fidelity Scene Editing

FlowR: Flowing from Sparse to Dense 3D Reconstructions

Hardware-Rasterized Ray-Based Gaussian Splatting

Easy3D: A Simple Yet Effective Method for 3D Interactive Segmentation

Online Learning With Bayesian Classification Trees

Loss Max-Pooling for Semantic Image Segmentation

In-Place Activated BatchNorm for Memory-Optimized Training of DNNs

Learning Multi-Object Tracking and Segmentation From Automatic Annotations

Improving Panoptic Segmentation at All Scales

AutoRF: Learning 3D Object Radiance Fields From Single View Observations

Panoptic Lifting for 3D Scene Understanding With Neural Fields

OrienterNet: Visual Localization in 2D Public Maps With Neural Matching

DiffRF: Rendering-Guided 3D Radiance Field Diffusion

Deep Neural Decision Forests

The Mapillary Vistas Dataset for Semantic Understanding of Street Scenes

Disentangling Monocular 3D Object Detection

Are We Missing Confidence in Pseudo-LiDAR Methods for Monocular 3D Object Detection?

Mapillary Planet-Scale Depth Dataset

Towards Generalization Across Depth for Monocular 3D Object Detection

Improving Optical Flow on a Pyramid Level

Seamless Scene Segmentation

Volumetric Surfaces: Representing Fuzzy Geometries with Layered Meshes

A Recipe for Generating 3D Worlds from a Single Image

Generative Gaussian Splatting: Generating 3D Scenes with Video Diffusion Priors

HybridNeRF: Efficient Neural Rendering via Adaptive Volumetric Surfaces

MultiDiff: Consistent Novel View Synthesis from a Single Image

Dropout distillation