Instructors: Prof. Dr. Nassir Navab, Dr. Shahrooz Faghih Roohi, Ashkan Khakzar, Azade Farshad, Yousef Yeganeh
Time: TBA
Registration
- Registration must be done through TUM Matching Platform (please pay attention to the Deadlines)
- In order to increase your priority, please also apply via our own Registration system.
- The maximum number of participants: 20.
Announcements
- It is highly recommended to attend the seminar in person. However, for those who cannot attend in person, the online link via Zoom will be provided (the meeting link will be shared with participants via email and on the moodle).
- The presentation and blogpost guidelines are available here: Guide_DLMA WS2022_23.pdf
Introduction
- Deep Learning is growing tremendously in Computer Vision and Medical Imaging as well. Highly impacted journals in the medical imaging community, i.e. IEEE Transaction on Medical Imaging, published recently their special edition on Deep Learning [1]. The Seminar will propose a list of recent scientific articles related to the main current research topics in deep learning for Medical Applications together with some interesting papers from other communities (CVPR, NeurIPS, ICCV, ICLR, ICML, ...).
Course Structure
In this Master Seminar (Hauptseminar), students select one scientific topic from the list provided by course organizers. The students should read the proposed sample papers by the tutors, find the topic-related articles, summarize and compare them in their presentation and blogpost:
- Presentation: The selected paper is presented to the other participants (Maximum 25 minutes presentation, 10 minutes questions). You can use the CAMP templates for PowerPoint TUM-Template.pptx.
- Blog Post: A blog post of 2000-2500 words, excluding references should be submitted before the deadline. The blog post must include all references used and must be written completely in your own words. Copy and paste will not be tolerated
- Attendance: Participants have to participate actively in all seminar sessions. Each presentation is followed by a discussion and everyone is encouraged to participate actively.
Submission Deadline: You have to submit the first draft of the blog post one week before your presentation session. However, you would have time to modify it until the last session. You should also submit the presentation one day right after your presentation session.
Schedule
TBA
| Date | Session: Topic | Slides | Students |
|---|---|---|---|
| 21/07 | Preliminary Meeting | ||
| 08/12 | Uncertainty estimation for medical segmentation | Yao, Junting | |
| 15/12 | Motion Compensation for Electromagnetically Tracked Catheters Transformers in Image Registration | Wally, Youssef Ünlü Idil | |
| 22/12 | Self-supervised Vessel Segmentation Automated Semantic/Anatomical Vessel labeling Masking Techniques for Computer Vision and Medical Imaging | Avendano-Prieto, Natalia Otto, Julia Anil Kul | |
| 12/01 | Representation Learning using Generative Models Graph / Scene Graph representation learning Incremental Learning in Medical Image Analysis: | Sun Mei Bigom, Andreas Dozono, Kohei | |
| 19/01 | Curvilinear structures segmentation Unsupervised Domain Adaptation in Medical Image Analysis SSL representation learning by/for object detection | Sotela, Sebastian Posada Cárdenas, Andrea Jiajun Wang | |
| 26/01 | Cross-modal learning Contrastive Language-Image Models in Medical Image Understanding Multimodal Image Synthesis | Duman, Serdar Diery, Kristina Karoui, Yasmine | |
| 02/02 | Diffusion Models for Medical Imaging Physics Informed Neural Network for Computer Vision and Medical Imaging NeRF Applications in Medical Imaging Self-Supervised/Unsupervised Image and Medical Imaging Segmentation | Zhang, Yifan He, Xinyi Matiunina Daria Palmisano Patrizio |
List of Topics and Material
Papers in this course are selected from the following venues/publications:
CVPR: Conference on Computer Vision and Pattern Recognition
ICLR: International Conference on Learning Representations
NeurIPS: Neural Information Processing Systems
TPAMI: IEEE Transactions on Pattern Analysis and Machine Intelligence
TMI: IEEE Transaction on Medical Imaging
JBHI: IEEE Journal of Biomedical and Health Informatics
MedIA: Medical Image Analysis (Elsevier)
MICCAI: Medical Image Computing and Computer-Assisted Intervention
BMVC: British Machine Vision Conference
MIDL: Medical Imaging with Deep Learning
List of topics
| No | Topic | Sample Papers | Journal/ Conference | Tutor | Student | Link |
|---|---|---|---|---|---|---|
| 1 | Learning-based Statistical Shape Model | Deep implicit statistical shape models for 3d medical image delineation | AAAI 2022 | None | https://ojs.aaai.org/index.php/AAAI/article/view/20110 | |
| Deep Structural Causal Shape Models | ECCV 2022 | https://arxiv.org/abs/2208.10950 | ||||
| Leveraging unsupervised image registration for discovery of landmark shape descriptor | MedIA 2021 | https://www.sciencedirect.com/science/article/abs/pii/S1361841521002036 | ||||
| 2 | Representation Learning using Generative Models | Generative Flow Networks for Discrete Probabilistic Modeling | ICML 2022 Spotlight | Sun Mei | https://proceedings.mlr.press/v162/zhang22v.html | |
| Flamingo: a Visual Language Model for Few-Shot Learning | DeepMind | https://arxiv.org/pdf/2204.14198.pdf | ||||
| Object Scene Representation Transformer | NeurIPS 2022 | https://arxiv.org/pdf/2206.06922.pdf | ||||
| 3 | Graph / Scene Graph representation learning | VARSCENE: A Deep Generative Model for Realistic Scene Graph Synthesis | ICML 2022 Spotlight | Bigom, Andreas | https://proceedings.mlr.press/v162/verma22b/verma22b.pdf | |
| Self-Supervised Representation Learning via Latent Graph Prediction | ICML 2022 Spotlight | https://proceedings.mlr.press/v162/xie22e.html | ||||
| Iterative Scene Graph Generation | NeurIPS 2022 | https://arxiv.org/pdf/2207.13440.pdf | ||||
| 4 | Motion Compensation for Electromagnetically Tracked Catheters | A survey of catheter tracking concepts and methodologies | Medical Image Analysis | Wally, Youssef | https://doi.org/10.1016/j.media.2022.102584 | |
| Respiratory motion compensation with tracked internal and external sensors during CT-guided procedures | Computer Aided Surgery | https://doi.org/10.3109/10929080600740871 | ||||
| Position Control of Motion Compensation Cardiac Catheters | IEEE Transactions on Robotics | https://doi.org/10.1109/TRO.2011.2160467 | ||||
| 5 | Diffusion Models for Medical Imaging | Diffusion Deformable Model for 4D Temporal Medical Image Generation | MICCAI 2022 | Zhang, Yifan | https://arxiv.org/pdf/2206.13295.pdf | |
| Diffusion Models for Medical Anomaly Detection | MICCAI 2022 | http://arxiv.org/abs/2203.04306 | ||||
| Fast Unsupervised Brain Anomaly Detection and Segmentation with Diffusion Models | MICCAI 2022 | http://arxiv.org/abs/2206.03461 | ||||
| 6 | Symbolic Regression | Discovering Symbolic Models from Deep Learning with Inductive Biases | NeurIPS 2020 | None | https://proceedings.neurips.cc/paper/2020/hash/c9f2f917078bd2db12f23c3b413d9cba-Abstract.html | |
| End-to-end symbolic regression with transformers | ArXiv 2022 | https://arxiv.org/pdf/2204.10532.pdf | ||||
| AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularity | NeurIPS 2020 | https://proceedings.neurips.cc/paper/2020/hash/33a854e247155d590883b93bca53848a-Abstract.html | ||||
| 7 | Uncertainty estimation for medical segmentation | Layer Ensembles: A Single-Pass Uncertainty Estimation in Deep Learning for Segmentation | MICCAI 2022 | Yao, Junting | https://arxiv.org/pdf/2203.08878.pdf | |
| Efficient Bayesian Uncertainty Estimation for nnU-Net | MICCAI 2022 | https://link.springer.com/chapter/10.1007/978-3-031-16452-1_51 | ||||
| CRISP - Reliable Uncertainty Estimation for Medical Image Segmentation | MICCAI 2022 | https://arxiv.org/abs/2206.07664 | ||||
| 8 | Cross-modal learning | Learning Explicit and Implicit Dual Common Subspaces for Audio-Visual Cross-Modal Retrieval | TOMM 2022 | Duman, Serdar | https://dl.acm.org/doi/pdf/10.1145/3564608 | |
VISUALVOICE: Audio-Visual Speech Separation with Cross-Modal Consistency (2021) | CVPR 2021 | https://arxiv.org/pdf/2101.03149.pdf | ||||
| Visually Guided Sound Source Separation and Localization Using Self-Supervised Motion Representations | IEEE/CVF/WACV 2022 | |||||
| 9 | Curvilinear structures segmentation | CS2-Net: Deep learning segmentation of curvilinear structures in medical imaging | MedIA 2021 | Sotela, Sebastian | https://www.sciencedirect.com/science/article/abs/pii/S1361841520302383 | |
| 3D vessel-like structure segmentation in medical images by an edge-reinforced network | MedIA 2022 | https://www.sciencedirect.com/science/article/abs/pii/S1361841522002201 | ||||
| 10 | Transformers in Image Registration | Deformer: Towards Displacement Field Learning for Unsupervised Medical Image Registration | MICCAI 2022 | Ünlü Idil | https://link.springer.com/chapter/10.1007/978-3-031-16446-0_14 | |
| TransMorph: Transformer for unsupervised medical image registration | MedIA 2022 | https://www.sciencedirect.com/science/article/abs/pii/S1361841522002432 | ||||
| Affine Medical Image Registration With Coarse-To-Fine Vision Transformer | CVPR 2022 | |||||
| 11 | Masking Techniques for Computer Vision and Medical Imaging | Resolution-robust large mask inpainting with fourier convolutions | WACV 2022 | Anil Kul | ||
| Image inpainting for irregular holes using partial convolutions | ECCV 2018 | |||||
| Incremental transformer structure enhanced image inpainting with masking positional encoding | CVPR 2022 | |||||
| 12 | Self-Supervised/Unsupervised Image and Medical Imaging Segmentation | Self-Supervised Learning for Few-Shot Medical Image Segmentation | TMI 2022 | Palmisano Patrizio | https://ieeexplore.ieee.org/abstract/document/9709261 | |
| Mix-and-match tuning for self-supervised semantic segmentation | AAAI 2018 | https://ojs.aaai.org/index.php/AAAI/article/view/12331 | ||||
| Unsupervised Image Segmentation by Mutual Information Maximization and Adversarial Regularization | RAL 2021 | https://ieeexplore.ieee.org/abstract/document/9476978 | ||||
| 13 | Physics Informed Neural Network for Computer Vision and Medical Imaging | Physics-Informed Neural Networks for Brain Hemodynamic Predictions Using Medical Imaging | TMI 2022 | He, Xinyi | https://ieeexplore.ieee.org/abstract/document/9740143 | |
| Physics-informed neural networks for myocardial perfusion MRI quantification | MedIA 2022 | https://www.sciencedirect.com/science/article/pii/S1361841522000512 | ||||
| Physics-Informed Neural Networks for Cardiac Activation Mapping | Frontiers 2020 | https://www.frontiersin.org/articles/10.3389/fphy.2020.00042/full | ||||
| 14 | NeRF Applications in Medical Imaging | ImplicitAtlas: Learning Deformable Shape Templates in Medical Imaging | CVPR 2022 | Matiunina Daria | ||
| 3D Ultrasound Spine Imaging with Application of Neural Radiance Field Method | IUS 2021 | https://ieeexplore.ieee.org/abstract/document/9593917 | ||||
| Implicit Neural Representations for Medical Imaging Segmentation | MICCAI 2022 | https://link.springer.com/chapter/10.1007/978-3-031-16443-9_42 | ||||
| 15 | Contrastive Language-Image Models in Medical Image Understanding | Contrastive Learning of Medical Visual Representations from Paired Images and Text | arxiv 2020/ revised 2022 | Diery, Kristina | https://arxiv.org/abs/2010.00747 | |
| RepsNet: Combining Vision with Language for Automated Medical Reports | MICCAI 2022 | https://link.springer.com/chapter/10.1007/978-3-031-16443-9_68 | ||||
| Radiological Reports Improve Pre-training for Localized Imaging Tasks on Chest X-Rays | MICCAI 2022 | https://link.springer.com/chapter/10.1007/978-3-031-16443-9_62 | ||||
| 16 | Multimodal Image Synthesis | Multimodal Image Synthesis and Editing: A Survey | PAMI 2022 | Karoui, Yasmine | https://arxiv.org/pdf/2112.13592.pdf | |
| Hi-net: hybrid-fusion network for multi-modal MR image synthesis | TMI 2020 | |||||
| ResViT: residual vision transformers for multimodal medical image synthesis | TMI 2022 | |||||
| 17 | Incremental Learning in Medical Image Analysis: | Continual Class Incremental Learning for CT Thoracic Segmentation | MICCAI 2020 | Dozono, Kohei | https://link.springer.com/chapter/10.1007/978-3-030-60548-3_11 | |
| Mnemonics Training: Multi-Class Incremental Learning without Forgetting | CVPR 2020 | |||||
| Class-Incremental Learning by Knowledge Distillationwith Adaptive Feature Consolidation | CVPR 2022 | |||||
| 18 | Unsupervised Domain Adaptation in Medical Image Analysis | S-CUDA: Self-cleansing unsupervised domain adaptation for medical image segmentation | MedIA 2021 | Posada Cárdenas, Andrea | https://www.sciencedirect.com/science/article/pii/S1361841521002590 | |
| Data Efficient Unsupervised Domain Adaptationfor Cross-Modality Image Segmentation | MICCAI 2019 | https://link.springer.com/chapter/10.1007/978-3-030-32245-8_74 | ||||
| Unsupervised Domain Adaptation for Medical Image Segmentation by Disentanglement Learning and Self-Training | TMI 2022 | |||||
| 19 | Self-supervised Vessel Segmentation | Self-supervised vessel segmentation via adversarial learning | ICCV2021 | Avendano-Prieto, Natalia | ||
| Diffusion Adversarial Representation Learning for Self-supervised Vessel Segmentation | Arxiv 2022 | https://arxiv.org/abs/2209.14566 | ||||
| 20 | Automated Semantic/Anatomical Vessel labeling | Automated anatomical labeling of a topologically variant abdominal arterial system via probabilistic hypergraph matching | MIA2022 | Otto, Julia | https://www.sciencedirect.com/science/article/abs/pii/S1361841521002942 | |
| CPR-GCN: Conditional Partial-Residual Graph Convolutional Network in Automated Anatomical Labeling of Coronary Arterie | CVPR2020 | |||||
| Automated anatomical labeling of coronary arteries via bidirectional tree LSTMs | IJCARS 2019 | https://link.springer.com/article/10.1007/s11548-018-1884-6 | ||||
| 21 | Out of Distribution (OOD) robustness | An impartial take to the cnn vs transformer robustness contest | ECCV 2022 | None | https://arxiv.org/abs/2207.11347 | |
| Plex: Towards Reliability using Pretrained Large Model Extensions | Arxiv 2022 (Google AI) | https://arxiv.org/abs/2207.07411 | ||||
| Smoothness in neural network approximators: the good, the bad, the ugly | NeurIPS 2020 (Deepmind) | |||||
| 22 | SSL representation learning by/for object detection | Object discovery and representation networks | ECCV 2022 | Jiajun Wang | https://arxiv.org/abs/2203.08777 | |
| Unsupervised part discovery from contrastive reconstruction | NeurIPS 2021 | https://arxiv.org/abs/2111.06349 | ||||
| Efficient Visual Pretraining with Contrastive Detection | Arxiv | https://arxiv.org/abs/2103.10957 | ||||
| 23 | Continual Learning | Architecture Matters in Continual Learning | Arxiv 2022 (Deepmind) | None | https://arxiv.org/abs/2202.00275 | |
| Efficient Continual Learning in Neural Network Subspaces | Arxiv 2022 (Deepmind) | https://arxiv.org/abs/2202.09826 | ||||
| Wide Neural Networks Forget Less Catastrophically | Arxiv 2022 (Deepmind) | https://arxiv.org/abs/2110.11526 |
Literature and Helpful Links
A lot of scientific publications can be found online.
The following list may help you to find some further information on your particular topic:
- Microsoft Academic Search
- Google Scholar
- CiteSeer
- CiteULike
- Collection of Computer Science Bibliographies
Some publishers:
- ScienceDirect (Elsevier Journals)
- IEEE Journals
- ACM Digital Library
Libraries (online and offline):
- http://rzblx1.uni-regensburg.de/ezeit/ (Elektronische Zeitschriften Bibliothek)
- Verbundkatalog des Bibliotheksverbundes Bayern (BVB)
- Computer ORG
- http://www.ub.tum.de/ (TUM Library)
- To get access onto the electronic library, see http://www.ub.tum.de/medien/ejournals/readme.html
- "proxy.biblio.tu-muenchen.de" mit Port 8080 (nur fuer http). Damit klappen zumindest portal.acm.org und computer.org meistens
- Various proceedings of conferences in our AR-Lab, 03.13.036 (These proceedings are not for lending!)
Some further hints for working with references:
- JabRef is a Java program for comfortable working with Bibtex literature databases. Handy feature: if you know the PubMed ID for an article, JabRef can import data from there (via "Web Search/Medline").
- Mendeley is a cross-platform program for organising your references.
If you find useful resources that are not already listed here, please tell us, so we can add them for others. Thanks.