Universität Ulm
Institute of Media Informatics
Visual Computing Group logo Visual Computing Group Ulm University

Pedro Hermosilla Casajus obtained his PhD in Computing from the Universitat Politecnica de Catalunya in 2017. In September 2017 he joined the Visual Computing group in Ulm as a PostDoc, where he is currently working on Point Cloud processing and Machine Learning techniques applied to Computer Vision and Computer Graphics problems. He has left the research group in Dezember 2022 to persue his accademic carrier at the Institute of Visual Computing and Human-Centered Technology, TU Vienna (2022).

Research interests

  • Point Cloud processing
  • Machine Learning
  • Computer Graphics

Lectures

  • GPU Programming

Open Topics for Theses

I offer different thesis topics related to Machine Learning applied to 3D data. If you are interested in writing a thesis on one of these topics, write me an email or drop by my office.

Publications

45 papers
Unified Semantic Transformer for 3D Scene Understanding
Unified Semantic Transformer for 3D Scene Understanding
Sebastian Koch, Johanna Wald, Hidenobu Matsuki, Pedro Hermosilla, Timo Ropinski, Federico Tombari
TMLR 2026 ↓PDF
S2D: Sparse-To-Dense Keymask Distillation for Unsupervised Video Instance Segmentation
S2D: Sparse-To-Dense Keymask Distillation for Unsupervised Video Instance Segmentation
Leon Sick, Lukas Hoyer, Dominik Engel, Pedro Hermosilla, Timo Ropinski
arXiv ‹›Code
UNITE - Unified Semantic Transformer for 3D Scene Understanding
UNITE - Unified Semantic Transformer for 3D Scene Understanding
Sebastian Koch, Johanna Wald, Hidenobu Matsuki, Pedro Hermosilla, Timo Ropinski, Federico Tombari
arXiv ↓PDF
OpenHype: Hyperbolic Embeddings for Hierarchical Open-Vocabulary Radiance Fields
OpenHype: Hyperbolic Embeddings for Hierarchical Open-Vocabulary Radiance Fields
Lisa Weijler, Sebastian Koch, Fabio Poiesi, Timo Ropinski, Pedro Hermosilla
NeurIPS 2025 ↓PDF
From Natural to Nanoscale: Training ControlNet on Scarce FIB-SEM Data for Augmenting Semantic Segmentation Data
From Natural to Nanoscale: Training ControlNet on Scarce FIB-SEM Data for Augmenting Semantic Segmentation Data
Hannah Kniesel, Pascal Rapp, Pedro Hermosilla, Timo Ropinski
ICCVW 2025 ↓PDF ‹›Code
DeepEM Playgound: Bringing Deep Learning to Electron Microscopy Labs
DeepEM Playgound: Bringing Deep Learning to Electron Microscopy Labs
Hannah Kniesel, Poonam Poonam, Tristan Payer, Tim Bergner, Pedro Hermosilla, Timo Ropinski
jmi 2025 ‹›Code
Masked Scene Modeling: Narrowing the Gap Between Supervised and Self-Supervised Learning in 3D Scene Understanding
Masked Scene Modeling: Narrowing the Gap Between Supervised and Self-Supervised Learning in 3D Scene Understanding
Pedro Hermosilla, Christian Stippel, Leon Sick
CVPR 2025 ‹›Code
Active Learning Inspired ControlNet Guidance for Augmenting Semantic Segmentation Datasets
Active Learning Inspired ControlNet Guidance for Augmenting Semantic Segmentation Datasets
Hannah Kniesel, Pedro Hermosilla, Timo Ropinski
arXiv ↓PDF
RelationField: Relate Anything in Radiance Fields
RelationField: Relate Anything in Radiance Fields
Sebastian Koch, Johanna Wald, Mirco Colosi, Narunas Vaskevicius, Pedro Hermosilla, Federico Tombari, Timo Ropinski
CVPR 2025 ↓PDF
CutS3D: Cutting Semantics in 3D for 2D Unsupervised Instance Segmentation
CutS3D: Cutting Semantics in 3D for 2D Unsupervised Instance Segmentation
Leon Sick, Dominik Engel, Sebastian Hartwig, Pedro Hermosilla, Timo Ropinski
ICCV 2025 ▶Video
Attention-Guided Masked Autoencoders For Learning Image Representations
Attention-Guided Masked Autoencoders For Learning Image Representations
Leon Sick, Dominik Engel, Pedro Hermosilla, Timo Ropinski
WACV 2025
HPSCAN: Human Perception-Based Scattered Data Clustering
HPSCAN: Human Perception-Based Scattered Data Clustering
Sebastian Hartwig, Christian van Onzenoodt, Dominik Engel, Pedro Hermosilla, Timo Ropinski
CGF 2025 ↓PDF
Open3DSG: Open-Vocabulary 3D Scene Graphs from Point Clouds with Queryable Objects and Open-Set Relationships
Open3DSG: Open-Vocabulary 3D Scene Graphs from Point Clouds with Queryable Objects and Open-Set Relationships
Sebastian Koch, Narunas Vaskevicius, Mirco Colosi, Pedro Hermosilla, Timo Ropinski
CVPR 2024 ↓PDF
Weakly Supervised Virus Capsid Detection with Image-Level Annotations in Electron Microscopy Images
Weakly Supervised Virus Capsid Detection with Image-Level Annotations in Electron Microscopy Images
Hannah Kniesel, Leon Sick, Tristan Payer, Tim Bergner, Kavitha Shaga Devan, Clarissa Read, Paul Walther, Timo Ropinski, Pedro Hermosilla
ICLR 2024 ↓PDF ‹›Code
Lang3DSG: Language-based contrastive pre-training for 3D scene graph prediction
Lang3DSG: Language-based contrastive pre-training for 3D scene graph prediction
Sebastian Koch, Pedro Hermosilla, Narunas Vaskevicius, Mirco Colosi, Timo Ropinski
3DV 2024
SGRec3D: Self-Supervised Pre-training for 3D Scene Graph prediction
SGRec3D: Self-Supervised Pre-training for 3D Scene Graph prediction
Sebastian Koch, Pedro Hermosilla, Narunas Vaskevicius, Mirco Colosi, Timo Ropinski
WACV 2024 ↓PDF
Unsupervised Semantic Segmentation Through Depth-Informed Feature Distillation and Sampling
Unsupervised Semantic Segmentation Through Depth-Informed Feature Distillation and Sampling
Leon Sick, Dominik Engel, Pedro Hermosilla, Timo Ropinski
CVPR 2024 ↓PDF ‹›Code ▶Video
Weakly-Supervised Optical Flow Estimation for Time-of-Flight
Weakly-Supervised Optical Flow Estimation for Time-of-Flight
Michael Schelling, Pedro Hermosilla, Timo Ropinski
WACV 2023 ↓PDF ‹›Code
Variance-Aware Weight Initialization for Point Convolutional Neural Networks
Variance-Aware Weight Initialization for Point Convolutional Neural Networks
Pedro Hermosilla, Michael Schelling, Tobias Ritschel, Timo Ropinski
ECCV 2022 ↓PDF ‹›Code
Learning Human Viewpoint Preferences from Sparsely Annotated Models
Learning Human Viewpoint Preferences from Sparsely Annotated Models
Sebastian Hartwig, Michael Schelling, Christian van Onzenoodt, Pere-Pau Vázquez, Pedro Hermosilla, Timo Ropinski
CGF 2022 ↓PDF ‹›Code ▶Video
Gaussian Mixture Convolution Networks
Gaussian Mixture Convolution Networks
Adam Celarek, Pedro Hermosilla, Bernhard Kerbl, Timo Ropinski, Michael Wimmer
ICLR 2022 ↓PDF
Clean Implicit 3D Structure from Noisy 2D STEM Images
Clean Implicit 3D Structure from Noisy 2D STEM Images
Hannah Kniesel, Timo Ropinski, Tim Bergner, Kavitha Shaga Devan, Clarissa Read, Paul Walther, Tobias Ritschel, Pedro Hermosilla
CVPR 2022 ↓PDF ‹›Code
RADU: Ray-Aligned Depth Update Convolutions for ToF Data Denoising
RADU: Ray-Aligned Depth Update Convolutions for ToF Data Denoising
Michael Schelling, Pedro Hermosilla, Timo Ropinski
CVPR 2022 ↓PDF ‹›Code ▶Video
Data-Driven Deep Density Estimation
Data-Driven Deep Density Estimation
Patrik Puchert, Pedro Hermosilla, Tobias Ritschel, Timo Ropinski
NCA 2021 ↓PDF ‹›Code
Intrinsic-Extrinsic Convolution and Pooling for Learning on 3D Protein Structures
Intrinsic-Extrinsic Convolution and Pooling for Learning on 3D Protein Structures
Pedro Hermosilla, Marco Schäfer, Matej Lang, Gloria Fackelmann, Pere-Pau Vázquez, Barbora Kozlíková, Michael Krone, Tobias Ritschel, Timo Ropinski
ICLR 2021 ‹›Code
Enabling Viewpoint Learning through Dynamic Label Generation
Enabling Viewpoint Learning through Dynamic Label Generation
Michael Schelling, Pedro Hermosilla, Pere-Pau Vázquez, Timo Ropinski
Eurographics 2021 ↓PDF ‹›Code
Teaser
GPU Pro: Advanced Rendering Techniques
Pedro Hermosilla, Pere-Pau Vázquez
Real-Time Visualization of 3D Amyloid-Beta Fibrils from 2D Cryo-EM Density Maps
Real-Time Visualization of 3D Amyloid-Beta Fibrils from 2D Cryo-EM Density Maps
Hannah Kniesel, Timo Ropinski, Pedro Hermosilla
EG VCBM 2020 ↓PDF
Void Space Surfaces to Convey Depth in Vessel Visualizations
Void Space Surfaces to Convey Depth in Vessel Visualizations
Julian Kreiser, Pedro Hermosilla, Timo Ropinski
IEEE VIS 2020 ↓PDF
Deep-learning the Latent Space of Light Transport
Deep-learning the Latent Space of Light Transport
Pedro Hermosilla, Sebastian Maisch, Tobias Ritschel, Timo Ropinski
CGF 2019 ↓PDF ▶Video
Total Denoising: Unsupervised Learning of 3D Point Cloud Cleaning
Total Denoising: Unsupervised Learning of 3D Point Cloud Cleaning
Pedro Hermosilla, Tobias Ritschel, Timo Ropinski
ICCV 2019 ↓PDF ‹›Code ▶Video
Teaser
A General Illumination Model for Molecular Visualization
Pedro Hermosilla, Pere-Pau Vázquez, Àlvar Vinacua, Timo Ropinski
CGF 2018
Teaser
Improving Perception of Molecular Surface Visualizations by Incorporating Translucency Effects
Pedro Hermosilla, Sebastian Maisch, Pere-Pau Vázquez, Timo Ropinski
EG VCBM 2018 ↓PDF
Monte Carlo Convolution for Learning on Non-Uniformly Sampled Point Clouds
Monte Carlo Convolution for Learning on Non-Uniformly Sampled Point Clouds
Pedro Hermosilla, Tobias Ritschel, Pere-Pau Vázquez, Àlvar Vinacua, Timo Ropinski
TOG 2018 ↓PDF ‹›Code
Teaser
Visual Analysis of Protein-Ligand Interactions
Pere-Pau Vázquez, Pedro Hermosilla, Victor Guallar, Jorge Estrada, Àlvar Vinacua
CGF 2018
Teaser
Visualization of Large Molecular Trajectories
David Duran Rosich, Pedro Hermosilla, Timo Ropinski, Barbora Kozlíková, Àlvar Vinacua, Pere-Pau Vázquez
IEEE VIS 2018
Teaser
Interactive GPU-based Generation of Solvent Excluded Surfaces
Pedro Hermosilla, Michael Krone, Victor Guallar, Pere-Pau Vázquez, Àlvar Vinacua, Timo Ropinski
Vis Comput 2017
Teaser
PhD Thesis: Advanced inspection techniques for molecular simulations
Pedro Hermosilla
Teaser
Physics-based Visual Characterization of Molecular Interaction Forces
Pedro Hermosilla, Jorge Estrada, Victor Guallar, Timo Ropinski, Àlvar Vinacua, Pere-Pau Vázquez
IEEE VIS 2017
Teaser
High quality illustrative effects for molecular rendering
Pedro Hermosilla, Victor Guallar, Àlvar Vinacua, Pere-Pau Vázquez
CAD and Graphics 2016
Teaser
Interactive inspection of complex multi-object industrial assemblies
Oscar Argudo, Isaac Besora, Pere Brunet, Carles Creus, Pedro Hermosilla, Isabel Navazo, Àlvar Vinacua
CAD 2016
Teaser
Adaptive on-the-fly molecular ribbons generation
Pedro Hermosilla, Victor Guallar, Àlvar Vinacua, Pere-Pau Vázquez
Teaser
Instant Visualization of Secondary Structures of Molecular Models
Pedro Hermosilla, Victor Guallar, Àlvar Vinacua, Pere-Pau Vázquez
EG VCBM 2015
Teaser
Uncertainty Visualization of Brain Fibers
Pedro Hermosilla, Ralph Brecheisen, Pere-Pau Vázquez, Anna Vilanova
CEIG 2012
Teaser
Single pass GPU stylized edges
Pedro Hermosilla, Pere-Pau Vázquez
SIAG 2009
← Back to Team