Publications

SeMoLi: What Moves Together Belongs Together

Unifying Short and Long-Term Tracking with Graph Hierarchies

Novel Class Discovery for 3D Point Cloud Semantic Segmentation

Learning to Discover and Detect Objects

Opening up Open-World Tracking

Low-budget label query through domain alignment enforcement

Forecasting from LiDAR via Future Object Detection

(Just) A Spoonful of Refinements Helps the Registration Error Go Down

Patch2Pix: Epipolar-Guided Pixel-Level Correspondences

EagerMOT: 3D Multi-Object Tracking via Sensor Fusion

4D Panoptic LiDAR Segmentation

HistoNet: Predicting size histograms of object instances

Single-shot Panoptic Segmentation

Make One-Shot Video Object Segmentation Efficient Again

Learning a Neural Solver for Multiple Object Tracking

How To Train Your Deep Multi-Object Tracker

4D Generic Video Object Proposals

Tracking without bells and whistles

Towards Generalizing Sensorimotor Control Across Weather Conditions

Deep Appearance Maps

Incremental learning for the detection and classification of gan-generated images

Video Object Segmentation Without Temporal Information

Fusion of Head and Full-Body Detectors for Multi-Object Tracking

Deep Depth from Focus

Combined Image- and World-Space Tracking in Traffic Scenes

One-Shot Video Object Segmentation

Learning by tracking: siamese CNN for robust target association