Publications

Recent work showed that large diffusion models can be reused as highly precise monocular depth estimators by casting depth estimation as an image-conditional image generation task. While the proposed model achieved state-of-the-art results, high computational demands due to multi-step inference limited its use in many scenarios. In this paper, we show that the perceived inefficiency was caused by a flaw in the inference pipeline that has so far gone unnoticed. The fixed model performs comparably to the best previously reported configuration while being more than 200x faster. To optimize for downstream task performance, we perform end-to-end fine-tuning on top of the single-step model with task-specific losses and get a deterministic model that outperforms all other diffusion-based depth and normal estimation models on common zero-shot benchmarks. We surprisingly find that this fine-tuning protocol also works directly on Stable Diffusion and achieves comparable performance to current state-of-the-art diffusion-based depth and normal estimation models, calling into question some of the conclusions drawn from prior works.

Interactive segmentation has an important role in facilitating the annotation process of future LiDAR datasets. Existing approaches sequentially segment individual objects at each LiDAR scan, repeating the process throughout the entire sequence, which is redundant and ineffective. In this work, we propose interactive 4D segmentation, a new paradigm that allows segmenting multiple objects on multiple LiDAR scans simultaneously, and Interactive4D, the first interactive 4D segmentation model that segments multiple objects on superimposed consecutive LiDAR scans in a single iteration by utilizing the sequential nature of LiDAR data. While performing interactive segmentation, our model leverages the entire space-time volume, leading to more efficient segmentation. Operating on the 4D volume, it directly provides consistent instance IDs over time and also simplifies tracking annotations. Moreover, we show that click simulations are crucial for successful model training on LiDAR point clouds. To this end, we design a click simulation strategy that is better suited for the characteristics of LiDAR data. To demonstrate its accuracy and effectiveness, we evaluate Interactive4D on multiple LiDAR datasets, where Interactive4D achieves a new state-of-the-art by a large margin.

Authors: Stephanie Käs, Sven Peter, Henrik Thillmann, Anton Burenko, Timm Linder, David Adrian, and Dennis Mack, Bastian Leibe

In this work, we tackle the challenge of 3D human pose estimation in fisheye images, which is crucial for applications in robotics, human-robot interaction, and automotive perception. Fisheye cameras offer a wider field of view, but their distortions make pose estimation difficult. We systematically analyze how different camera models impact prediction accuracy and introduce a strategy to improve pose estimation across diverse viewing conditions.

A key contribution of our work is FISHnCHIPS, a novel dataset featuring 3D human skeleton annotations in fisheye images, including extreme close-ups, ground-mounted cameras, and wide-FOV human poses. To support future research, we will be publicly releasing this dataset.

More details coming soon — stay tuned for the final publication! Looking forward to sharing our findings at ICRA 2025!