Pedro Hermosilla Casajus
Assistant Prof. Doctor
Research Areas
Research Topics
- Computer Vision, Artificial intelligence, 3D Computer Vision, Machine Learning
About
My research is focused on developing machine learning technologies for 3D and unstructured data, with a special interest in point clouds, graphs, and implicit representations. During my research, I applied these technologies to address various problems in Computer Vision, Computer Graphics, and Bioinformatics.
Role
- Assistant Professor
Computer Vision, E193-01
Contact
- pedro.hermosilla@tuwien.ac.at
- +43-1-58801-193157
- 1040 Wien, Favoritenstrasse 11 / Room HE0434
- vCard from TISS
Courses
Summer 2027
- Project Media and Human-Centered Computing 1 / 193.021 PR
- Project Media and Human-Centered Computing 2 / 193.022 PR
Winter 2026
- Bachelor Thesis / 183.582 PR
- Doctoral Seminar / 183.619 SE
- Fundamentals of Computer Vision / 193.125 VU
- Methods for Data Generation and Analytics in Medicine and Life Sciences / 193.026 VU
- Project in Computer Science 1 / 193.117 PR
- Project in Computer Science 2 / 193.118 PR
- Project in Medical Informatics / 193.025 PR
- Project in Visual Computing 1 / 193.023 PR
- Project in Visual Computing 2 / 193.024 PR
- Project Media and Human-Centered Computing 1 / 193.021 PR
- Project Media and Human-Centered Computing 2 / 193.022 PR
- The Essentials of AI / 192.216 VU
Projects
- 2026 – 2028 / Austrian Research Promotion Agency (FFG)
- 2025 – 2028 / Austrian Research Promotion Agency (FFG) / Publication
- 2026 – 2027 / Magna Steyr Fahrzeugtechnik GmbH &
- 2026 – 2027 / Kapsch
- 2024 – 2026 / cirqular pointcloud analytics GmbH
Publications
- Kovács, Á. S., Hermosilla, P., Raidou, R. G. (2026). Style Brush: Guided Style Transfer for 3D Objects. Computer Graphics Forum, Article e70308. Peer-reviewed.DOI: 10.1111/cgf.70308
- Grammatikaki, A., Eschner, J., Hermosilla, P., Argudo, O., Waldner, M. (2026). TreeON: Reconstructing 3D Tree Point Clouds from Orthophotos and Heightmaps. Computer Graphics Forum, Article e70366. Invited and peer-reviewed.
- Johannes Spöcklberger, Lin, W., Hermosilla, P., Doveh, S., Possegger, H., Mirza, J. M. (2025). Exploring Modality Guidance to Enhance VFM-Based Feature Fusion for UDA in 3D Semantic Segmentation. In 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) (pp. 4789–4798). IEEE. Peer-reviewed.
- Kniesel, H., Poonam, P., Payer, T., Bergner, T., Hermosilla, P., Ropinski, T. (2025). DeepEM Playground: Bringing deep learning to electron microscopy labs. Journal of Microscopy, 299(3), 287–300. Peer-reviewed.DOI: 10.1111/jmi.70005
- Weijler, L. M., Mirza, J. M., Sick, L., Ekkazan, C., Hermosilla, P. (2025). TTT-KD: Test-Time Training for 3D Semantic Segmentation Through Knowledge Distillation From Foundation Models. In 2025 International Conference on 3D Vision (3DV) (pp. 1264–1274). IEEE. Peer-reviewed.
- Koch, S., Wald, J., Colosi, M., Vaskevicius, N., Hermosilla, P., Tombari, F., Ropinski, T. (2025). RelationField: Relate Anything in Radiance Fields. In 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 21706–21716). IEEE. Peer-reviewed.
- Sick, L., Engel, D., Hermosilla, P., Ropinski, T. (2025). Attention-Guided Masked Autoencoders for Learning Image Representations. In 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) (pp. 836–846). IEEE. Peer-reviewed.
- Weijler, L., Reiter, M., Hermosilla, P., Maurer-Granofszky, M., Dworzak, M. (2025). On the Importance of Local and Global Feature Learning for Automated Measurable Residual Disease Detection in Flow Cytometry Data. In A. Antonacopoulos, S. Chaudhuri, R. Chellappa, Cl. Liu, S. Bhattacharya, U. Pal (Eds.), Pattern Recognition (pp. 316–331). Peer-reviewed.DOI: 10.1007/978-3-031-78198-8_21 / Project: MyeFLOW
- Weijler, L. M., Hermosilla Casajus, P. (2025). Efficient Continuous Group Convolutions for Local SE(3) Equivariance in 3D Point Clouds. In 2025 International Conference on 3D Vision (3DV) (pp. 1372–1381). IEEE. Peer-reviewed.
- Hermosilla, P., Stippel, C., Sick, L. (2025). Masked Scene Modeling: Narrowing the Gap Between Supervised and Self-Supervised Learning in 3D Scene Understanding. In Proceedings of the IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR) (pp. 14835–14844). Peer-reviewed.
- Sick, L., Engel, D., Hartwig, S., Hermosilla, P., Ropinski, T. (2025). CutS3D: Cutting Semantics in 3D for 2D Unsupervised Instance Segmentation. In 2025 IEEE/CVF International Conference on Computer Vision (ICCV) (pp. 21265–21275). IEEE. Peer-reviewed.
- Weijler, L. M., Koch, S., Poiesi, F., Ropinski, T., Hermosilla, P. (2025). OpenHype: Hyperbolic Embeddings for Hierarchical Open-Vocabulary Radiance Fields. In Advances in Neural Information Processing Systems 38 Main Conference : 39th Conference on Neural Information Processing Systems (NeurIPS 2025) (pp. 1–30). Neural Information Processing Systems Foundation, Inc. (NeurIPS). Peer-reviewed.DOI: 10.52202/085713-3972
- Hartwig, S., Onzenoodt, C. v., Engel, D., Hermosilla Casajus, P., Ropinski, T. (2025). HPSCAN: Human Perception‐Based Scattered Data Clustering. Computer Graphics Forum, 44(1), Article e15275. Peer-reviewed.DOI: 10.1111/cgf.15275
- Stippel, C., Mujkanovic, F., Leimkühler, T., Hermosilla Casajus, P. (2025). Marching Neurons: Accurate Surface Extraction for Neural Implicit Shapes. ACM Transactions on Graphics, 44(6), Article 222. Peer-reviewed.DOI: 10.1145/3763328
- Kniesel, H., Rapp, P., Hermosilla, P., Ropinski, T. (2025). From Natural to Nanoscale: Training ControlNet on Scarce FIB-SEM Data for Augmenting Semantic Segmentation Data. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops 2025 (pp. 5731–5740). Peer-reviewed.HDL: 20.500.12708/223672
- Kovacs, A. S., Hermosilla Casajus, P., Raidou, R. G. (2024). Surface-aware Mesh Texture Synthesis with Pre-trained 2D CNNs. Computer Graphics Forum, 43(2), Article e15016. Peer-reviewed.DOI: 10.1111/cgf.15016
- Erler, P., Fuentes‐Perez, L., Hermosilla, P., Guerrero, P., Pajarola, R., Wimmer, M. (2024). PPSurf: combining patches and point convolutions for detailed surface reconstruction. Computer Graphics Forum, Article e15000. Peer-reviewed.DOI: 10.1111/cgf.15000 / Download: PDF / Projects: Echtzeit-Formenakquisition mit sensorspezifischer Präzision, EVOCATION, Superhumans, Modeling the World at Scale
- Weijler, L., Kowarsch, F., Reiter, M., Hermosilla, P., Maurer-Granofszky, M., Dworzak, M. (2024). FATE: Feature-Agnostic Transformer-based Encoder for learning generalized embedding spaces in flow cytometry data. In 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) (pp. 7941–7949). Peer-reviewed.DOI: 10.1109/WACV57701.2024.00777 / Project: MyeFLOW
- Koch, S., Hermosilla, P., Vaskevicius, N., Colosi, M., Ropinski, T. (2024). Lang3DSG: Language-based contrastive pre-training for 3D Scene Graph prediction. In 2024 International Conference on 3D Vision (3DV) (pp. 1037–1047). Peer-reviewed.
- Sick, L., Engel, D., Hermosilla, P., Ropinski, T. (2024). Unsupervised Semantic Segmentation Through Depth-Guided Feature Correlation and Sampling. In 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 3637–3646). Peer-reviewed.
- Koch, S., Vaskevicius, N., Colosi, M., Hermosilla, P., Ropinski, T. (2024). Open3DSG: Open-Vocabulary 3D Scene Graphs from Point Clouds with Queryable Objects and Open-Set Relationships. In 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 14183–14193). Peer-reviewed.
- Koch, S., Hermosilla, P., Vaskevicius, N., Colosi, M., Ropinski, T. (2024). SGRec3D: Self-Supervised 3D Scene Graph Learning via Object-Level Scene Reconstruction. In 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) (pp. 3392–3402). Peer-reviewed.
- Kniesel, H., Sick, L., Payer, T., Bergner, T., Shaga Devan, K., Read, C., Walther, P., Ropinski, T., Hermosilla Casajus, P. (2024). Weakly Supervised Virus Capsid Detection with Image-Level Annotations in Electron Microscopy Images. In The Twelfth International Conference on Learning Representations. The Twelfth International Conference on Learning Representations (ICLR 2024), Wien, Austria. Peer-reviewed.HDL: 20.500.12708/203934
- Kovács, Á. S., Hermosilla, P., Raidou, R. G. (2024). G-Style: Stylized Gaussian Splatting. Computer Graphics Forum, 43(7), Article e15259. Peer-reviewed.DOI: 10.1111/cgf.15259
Presentations
- Weijler, L. M., Hermosilla, P. (2024, February 15). Efficient continuous group convolutions for local SO(3) equivariance in 3D point clouds [Conference Presentation]. COMPUTER VISION WINTER WORKSHOP 2024, Slovenia. Invited.HDL: 20.500.12708/203719
Theses
- Staats, M. (2024). Alpine terrain relighting : deep-learning based single image shadow-removal with digital elevation models [Diploma Thesis, Technische Universität Wien]. reposiTUm.DOI: 10.34726/hss.2024.112641 / Download: PDF
And more…
Soon, this page will include additional information such as reference projects, activities as journal reviewer and editor, memberships in councils and committees, and other research activities.
Until then, please visit Pedro’s research profile in TISS.