Explanipedia

From open-vocabulary to vocabulary-free semantic segmentation Open

Klara Reichard, Giulia Rizzoli, Stefano Gasperini, Lukas Hoyer, Pietro Zanuttigh , et al. · 2025

GaRA-SAM: Robustifying Segment Anything Model with Gated-Rank Adaptation Open

Yeho Gwon, Lukas Hoyer, Suha Kwak · 2025

Improving robustness of the Segment Anything Model (SAM) to input degradations is critical for its deployment in high-stakes applications such as autonomous driving and robotics. Our approach to this challenge prioritizes three key aspects…

From Open-Vocabulary to Vocabulary-Free Semantic Segmentation Open

Klara Reichard, Giulia Rizzoli, Stefano Gasperini, Lukas Hoyer, Pietro Zanuttigh , et al. · 2025

Open-vocabulary semantic segmentation enables models to identify novel object categories beyond their training data. While this flexibility represents a significant advancement, current approaches still rely on manually specified class nam…

MICDrop: Masking Image and Depth Features via Complementary Dropout for Domain-Adaptive Semantic Segmentation Open

Linyan Yang, Lukas Hoyer, Mark Weber, Tobias Fischer, Dengxin Dai , et al. · 2024

Unsupervised Domain Adaptation (UDA) is the task of bridging the domain gap between a labeled source domain, e.g., synthetic data, and an unlabeled target domain. We observe that current UDA methods show inferior results on fine structures…

Scribbles for All: Benchmarking Scribble Supervised Segmentation Across Datasets Open

Wolfgang Boettcher, Lukas Hoyer, Ozan Unal, Jan Eric Lenssen, Bernt Schiele · 2024

In this work, we introduce Scribbles for All, a label and training data generation algorithm for semantic segmentation trained on scribble labels. Training or fine-tuning semantic segmentation models with weak supervision has become an imp…

DGInStyle: Domain-Generalizable Semantic Segmentation with Image Diffusion Models and Stylized Semantic Control Open

Yuru Jia, Lukas Hoyer, Shengyu Huang, Tianfu Wang, Luc Van Gool , et al. · 2023

Large, pretrained latent diffusion models (LDMs) have demonstrated an extraordinary ability to generate creative content, specialize to user data through few-shot fine-tuning, and condition their output on other modalities, such as semanti…

SemiVL: Semi-Supervised Semantic Segmentation with Vision-Language Guidance Open

Lukas Hoyer, David Joseph Tan, Muhammad Ferjad Naeem, Luc Van Gool, Federico Tombari · 2023

In semi-supervised semantic segmentation, a model is trained with a limited number of labeled images along with a large corpus of unlabeled images to reduce the high annotation effort. While previous methods are able to learn good segmenta…

2D Feature Distillation for Weakly- and Semi-Supervised 3D Semantic Segmentation Open

Ozan Unal, Dengxin Dai, Lukas Hoyer, Yiğit Baran Can, Luc Van Gool · 2023

As 3D perception problems grow in popularity and the need for large-scale labeled datasets for LiDAR semantic segmentation increase, new methods arise that aim to reduce the necessity for dense annotations by employing weakly-supervised tr…

SILC: Improving Vision Language Pretraining with Self-Distillation Open

Muhammad Ferjad Naeem, Yongqin Xian, Xiaohua Zhai, Lukas Hoyer, Luc Van Gool , et al. · 2023

Image-Text pretraining on web-scale image caption datasets has become the default recipe for open vocabulary classification and retrieval models thanks to the success of CLIP and its variants. Several works have also used CLIP features for…

LiDAR Meta Depth Completion Open

Wolfgang Boettcher, Lukas Hoyer, Ozan Unal, Dengxin Dai · 2023

Depth estimation is one of the essential tasks to be addressed when creating mobile autonomous systems. While monocular depth estimation methods have improved in recent times, depth completion provides more accurate and reliable depth maps…

Improving Semi-Supervised and Domain-Adaptive Semantic Segmentation with Self-Supervised Depth Estimation Open

Lukas Hoyer, Dengxin Dai, Qin Wang, Yuhua Chen, Luc Van Gool · 2023

Training deep networks for semantic segmentation requires large amounts of labeled training data, which presents a major challenge in practice, as labeling segmentation masks is a highly labor-intensive process. To address this issue, we p…

EDAPS: Enhanced Domain-Adaptive Panoptic Segmentation Open

Suman Saha, Lukas Hoyer, Anton Obukhov, Dengxin Dai, Luc Van Gool · 2023

With autonomous industries on the rise, domain adaptation of the visual perception stack is an important research direction due to the cost savings promise. Much prior art was dedicated to domain-adaptive semantic segmentation in the synth…

Domain Adaptive and Generalizable Network Architectures and Training Strategies for Semantic Image Segmentation Open

Lukas Hoyer, Dengxin Dai, Luc Van Gool · 2023

Unsupervised domain adaptation (UDA) and domain generalization (DG) enable machine learning models trained on a source domain to perform well on unlabeled or even unseen target domains. As previous UDA&DG semantic segmentation methods are …

MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation Open

Lukas Hoyer, Dengxin Dai, Haoran Wang, Luc Van Gool · 2022

In unsupervised domain adaptation (UDA), a model trained on source data (e.g. synthetic) is adapted to target data (e.g. real-world) without access to target annotation. Most previous UDA methods struggle with classes that have a similar v…

HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation Open

Lukas Hoyer, Dengxin Dai, Luc Van Gool · 2022

Unsupervised domain adaptation (UDA) aims to adapt a model trained on the source domain (e.g. synthetic data) to the target domain (e.g. real-world data) without requiring further annotations on the target domain. This work focuses on UDA …

HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation Open

Lukas Hoyer, Dengxin Dai, Luc Van Gool · 2022

Unsupervised domain adaptation (UDA) aims to adapt a model trained on the source domain (e.g. synthetic data) to the target domain (e.g. real-world data) without requiring further annotations on the target domain. This work focuses on UDA …

DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentation Open

Lukas Hoyer, Dengxin Dai, Luc Van Gool · 2021

As acquiring pixel-wise annotations of real-world images for semantic segmentation is a costly process, a model can instead be trained with more accessible synthetic data and adapted to real images without requiring their annotations. This…

Improving Semi-Supervised and Domain-Adaptive Semantic Segmentation with Self-Supervised Depth Estimation Open

Lukas Hoyer, Dengxin Dai, Qin Wang, Yuhua Chen, Luc Van Gool · 2021

Training deep networks for semantic segmentation requires large amounts of labeled training data, which presents a major challenge in practice, as labeling segmentation masks is a highly labor-intensive process. To address this issue, we p…

Three Ways to Improve Semantic Segmentation with Self-Supervised Depth Estimation Open

Lukas Hoyer, Dengxin Dai, Yuhua Chen, Adrian Köring, Suman Saha , et al. · 2021

Training deep networks for semantic segmentation requires large amounts of labeled training data, which presents a major challenge in practice, as labeling segmentation masks is a highly labor-intensive process. To address this issue, we p…

Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation Open

Qin Wang, Dengxin Dai, Lukas Hoyer, Olga Fink, Luc Van Gool · 2021

Domain adaptation for semantic segmentation aims to improve the model performance in the presence of a distribution shift between source and target domain. Leveraging the supervision from auxiliary tasks~(such as depth estimation) has the …

Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation Open

Qin Wang, Dengxin Dai, Lukas Hoyer, Luc Van Gool, Olga Fink · 2021

Domain adaptation for semantic segmentation aims to improve the model performance in the presence of a distribution shift between source and target domain. Leveraging the supervision from auxiliary tasks~(such as depth estimation) has the …

Three Ways to Improve Semantic Segmentation with Self-Supervised Depth\n Estimation Open

Lukas Hoyer, Dengxin Dai, Yuhua Chen, Adrian Köring, Suman Saha , et al. · 2020

Training deep networks for semantic segmentation requires large amounts of\nlabeled training data, which presents a major challenge in practice, as\nlabeling segmentation masks is a highly labor-intensive process. To address\nthis issue, w…

Short-Term Prediction and Multi-Camera Fusion on Semantic Grids Open

Lukas Hoyer, Patrick Kesper, Anna Khoreva, Volker Fischer · 2019

An environment representation (ER) is a substantial part of every autonomous system. It introduces a common interface between perception and other system components, such as decision making, and allows downstream algorithms to deal with ab…

Grid Saliency for Context Explanations of Semantic Segmentation Open

Lukas Hoyer, Mauricio Muñoz, Prateek Katiyar, Anna Khoreva, Volker Fischer · 2019

Recently, there has been a growing interest in developing saliency methods that provide visual explanations of network predictions. Still, the usability of existing methods is limited to image classification models. To overcome this limita…

A Robot Localization Framework Using CNNs for Object Detection and Pose Estimation Open

Lukas Hoyer, Christoph Steup, Sanaz Mostaghim · 2018

External localization is an essential part for the indoor operation of small or cost-efficient robots, as they are used, for example, in swarm robotics. We introduce a two-stage localization and instance identification framework for arbitr…

A Robot Localization Framework Using CNNs for Object Detection and Pose\n Estimation Open

Lukas Hoyer, Christoph Steup, Sanaz Mostaghim · 2018

External localization is an essential part for the indoor operation of small\nor cost-efficient robots, as they are used, for example, in swarm robotics. We\nintroduce a two-stage localization and instance identification framework for\narb…

Lukas Hoyer YOU? Author Swipe