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
- The presentation and blogpost guidelines are available here: Guide_DLMA SS2022.pdf
- The preliminary meeting slides can be found here: DLMASS22.pdf
- The preliminary meeting is scheduled for Feb 3, 10:30 (Zoom link is visible on TUMonline).
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 actively participate.
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, July 28th. You should also submit the presentation one day right after your presentation session.
Schedule
TBA
| Date | Session: Topics | Slides | Students |
|---|---|---|---|
| 12.05 | Preliminary Meeting | ||
| 09.06 | Few-shot Image Synthesis Image-to-Image Translation | Juan Carlos Climent Pardo Wang, Yihao | |
| 23.06 | Vision Transformers Transformers for Segmentation Medical Visual Question Answering (VQA) | Demir, Ufuk Ganß, Marcel Demmel, Julia | |
| 30.06 | Semi/self-supervised Methods for Vessel Segmentation Task Modelling in Meta-learning Unsupervised Domain Adaptation for Segmentation | Hasny, Marta Chenyang Li Fabian Scherer | |
| 07.07 | Semi-Supervised Learning /Semi-Supervised Federated Learning Contrastive Learning/Trends in Self-Supervised Learning Unsupervised Anomaly Detection | Młynarczyk, Dominika Schreiber, Manuel Trotman, Rachelle | |
| 14.07 | Neural Network Robustness (adversarial examples) Neural Network Verification | Çelik, Furkan Engstler, Paul | |
| 21.07 | Shape-aware semi-supervised image segmentation Shape Completion Trends in Data Augmentation | Capelle, Finn Konov Mikhail Salah, Skander | |
| 28.07 | Deep Learning-based medical image registration Representation Learning using Generative Models implicit neural representations with deformation | Zhang, Zichen Bohosyan, Aleks Bou Orm, Ali |
List of Topics and Material
The proposed papers for each topic in this course are usually 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 papers
| No | Topic | Sample Papers | Journal/ Conference | Tutor | Student | Link |
|---|---|---|---|---|---|---|
| 1 | Shape-aware semi-supervised image segmentation | SimCVD: Simple Contrastive Voxel-Wise Representation Distillation for Semi-Supervised Medical Image Segmentation | TMI 2020 | Capelle, Finn | https://ieeexplore.ieee.org/abstract/document/9740182 | |
| 3D Graph-S2Net: Shape-Aware Self-ensembling Network for Semi-supervised Segmentation with Bilateral Graph Convolution | MICCAI 2021 | https://link.springer.com/chapter/10.1007/978-3-030-87196-3_39 | ||||
| 2 | Few-shot Image Synthesis | Towards faster and stabilized gan training for high-fidelity few-shot image synthesis | ICLR 2020 | Juan Carlos Climent Pardo | https://openreview.net/forum?id=1Fqg133qRaI | |
| One-Shot GAN: Learning To Generate Samples From Single Images and Videos | CVPR 2021 | |||||
| 3 | Semi/self-supervised Methods for Vessel Segmentation | Dual-consistency semi-supervision combined with self-supervision for vessel segmentation in retinal OCTA images | Biomedical Optics Express 2022 | Shahrooz Faghihroohi | Hasny, Marta | https://opg.optica.org/boe/fulltext.cfm?uri=boe-13-5-2824&id=471607 |
| LIFE: A Generalizable Autodidactic Pipeline for 3D OCT-A Vessel Segmentation | MICCAI 2021 | https://link.springer.com/chapter/10.1007/978-3-030-87193-2_49 | ||||
| 4 | Shape Completion | GRNet: Gridding Residual Network for Dense Point Cloud Completion | ECCV 2020 | Konov Mikhail | https://arxiv.org/pdf/2006.03761.pdf | |
| AutoSDF: Shape Priors for 3D Completion, Reconstruction and Generation | CVPR 2022 | https://arxiv.org/pdf/2203.09516.pdf | ||||
| 5 | Semi-Supervised Learning /Semi-Supervised Federated Learning | A Realistic Evaluation of Semi-Supervised Learning for Fine-Grained Classification | CVPR 2021 | Młynarczyk, Dominika | ||
| Semi-supervised Medical Image Segmentation via a Tripled-uncertainty Guided Mean Teacher Model with Contrastive Learning | MedIA 2022 | https://www.sciencedirect.com/science/article/abs/pii/S1361841522000925 | ||||
| 6 | Deep Learning based medical image registration | Orientation Estimation of Abdominal Ultrasound Images with Multi-Hypotheses Networks | MIDL 2022 | Zhang, Zichen | https://openreview.net/pdf?id=1gsauv2B7Ar | |
| Cross-Modal Attention for MRI and Ultrasound Volume Registration | MICCAI 2021 | https://link.springer.com/content/pdf/10.1007/978-3-030-87202-1_7.pdf | ||||
| 7 | Trends in Data Augmentation | Fast AdvProp | ICLR2022 | Yeganeh, Y. M. | Salah, Skander | https://openreview.net/forum?id=hcoswsDHNAW |
| SwapMix: Diagnosing and Regularizing the Over-Reliance on Visual Context in Visual Question Answering | CVPR2022 | https://arxiv.org/abs/2204.02285 | ||||
| 8 | Contrastive Learning/Trends in Self-Supervised Learning | VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning | NeurIPS 2021 | Schreiber, Manuel | https://arxiv.org/abs/2105.04906 | |
| Masked Siamese Networks for Label-Efficient Learning | CVPR2022 | https://arxiv.org/abs/2204.07141 | ||||
| 9 | Self-Supervised Learnig with False Negative Selection | SAM: Self-supervised Learning of Pixel-wise Anatomical Embeddings in Radiological Images | TMI 2021 | Tariq Bdair | Not Assigned | |
| Boosting Contrastive Self-Supervised Learning with False Negative Cancellation | WACV 2022 | |||||
| 10 | Unsupervised Anomaly Detection | Unsupervised Anomaly Detection in Medical Images with a Memory-augmented Multi-level Cross-attentional Masked Autoencoder | CVPR 2022 | Trotman, Rachelle | https://arxiv.org/pdf/2201.13271v1.pdf | |
| StRegA: Unsupervised Anomaly Detection in Brain MRIs using a Compact Context-encoding Variational Autoencoder | CVPR 2022 | https://arxiv.org/pdf/2203.11725v1.pdf | ||||
| 11 | Transformers for Segmentation | Unsupervised Anomaly Detection in Medical Images with a Memory-augmented Multi-level Cross-attentional Masked Autoencoder | ICCV2021 | Ganß, Marcel | https://arxiv.org/abs/2203.11725 | |
| StRegA: Unsupervised Anomaly Detection in Brain MRIs using a Compact Context-encoding Variational Autoencoder | CVPR2021 | https://arxiv.org/abs/2201.13271 | ||||
| 12 | Neural Network Robustness (adversarial examples) | Adversarial Examples Are Not Bugs, They Are Features | ICLR 2019 | Unbekannter Benutzer (ga59mat) | Çelik, Furkan | https://arxiv.org/abs/1905.02175 |
| Reading Race: AI Recognises Patient's Racial Identity In Medical Images | https://arxiv.org/abs/2107.10356 | |||||
| 13 | Neural Network Verification | Bias Field Robustness Verification of Large Neural Image Classifiers | BMVC 2021 | Unbekannter Benutzer (ga59mat) | Engstler, Paul | https://www.bmvc2021-virtualconference.com/assets/papers/1291.pdf |
| Efficient Neural Network Verification via Adaptive Refinement and Adversarial Search | ECAI 2020 | https://ebooks.iospress.nl/doi/10.3233/FAIA200385 | ||||
| 14 | Task Modelling in Meta-learning | Meta-Learning with Fewer Tasks through Task Interpolation | ICLR 2022 | Azade Farshad | Chenyang Li | https://arxiv.org/pdf/2106.02695.pdf |
| Task Relatedness-Based Generalization Bounds for Meta Learning | ICLR 2022 | https://openreview.net/pdf?id=A3HHaEdqAJL | ||||
| 15 | Knowledge Distillation in Meta-learning (Optional) | BERT Learns to Teach: Knowledge Distillation with Meta Learning | ACL 2022 | Azade Farshad | Not Assigned | https://arxiv.org/pdf/2106.04570.pdf |
| Online Hyperparameter Meta-Learning with Hypergradient Distillation | ICLR 2022 | https://arxiv.org/pdf/2110.02508.pdf | ||||
| 16 | Representation Learning using Generative Models | Shape your Space: A Gaussian Mixture Regularization Approach to Deterministic Autoencoders | NeurIPS 2021 | Azade Farshad | Bohosyan, Aleks | https://proceedings.neurips.cc/paper/2021/file/3c057cb2b41f22c0e740974d7a428918-Paper.pdf |
| GENERATIVE MODELS AS A DATA SOURCE FOR MULTIVIEW REPRESENTATION LEARNING | ICLR 2022 | https://arxiv.org/pdf/2106.05258.pdf | ||||
| 17 | Vision Transformers | An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale | ICLR 2021 | Yordanka Velikova | Demir, Ufuk | https://arxiv.org/pdf/2010.11929.pdf |
| Masked-attention Mask Transformer for Universal Image Segmentation | CVPR 2022 | https://arxiv.org/pdf/2112.01527v2.pdf | ||||
| 18 | Image-to-Image Translation | The Spatially-Correlative Loss for Various Image Translation Tasks (LSeSim ) | ICCV 2021 | Wang, Yihao | https://arxiv.org/pdf/2104.00854.pdf | |
| Contrastive learning for un-paired image-to-image translation | ECCV 2020 | https://arxiv.org/pdf/2007.15651.pdf | ||||
| 19 | Medical Visual Question Answering (VQA) | MMBERT: Multimodal BERT Pretraining for Improved Medical VQA | ISBI 2021 | Matthias Keicher | Demmel, Julia | |
| Vision-Language Transformer for Interpretable Pathology Visual Question Answering | IEEE Journal of Biomedical and Health Informatics | |||||
| 20 | implicit neural representations with deformation | Nerfies: Deformable Neural Radiance Fields | ICCV 2021 | Bou Orm, Ali | https://arxiv.org/pdf/2011.12948 | |
| D-NeRF: Neural Radiance Fields for Dynamic Scenes | CVPR 2021 | https://arxiv.org/abs/2011.13961 | ||||
| 21 | Unsupervised Domain Adaptation for Segmentation | Self-Attentive Spatial Adaptive Normalization for Cross-Modality Domain Adaptation | TMI 2021 | Fabian Scherer | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9380742 | |
| DANNet: A One-Stage Domain Adaptation Network for Unsupervised Nighttime Semantic Segmentation | CVPR 2021 | https://arxiv.org/pdf/2104.10834.pdf |
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.
2 Kommentare
Fabian Scherer sagt:
Hello, in the pdf about the structure of the Blog Post (https://wiki.tum.de/download/attachments/1046907253/Guide_DLMA%20SS2022.pdf?version=1&modificationDate=1653688942210&api=v2), something is broken. The sentence "It should summarize the topic with high emphasis on the" just stops without telling me on what I should emphasis
Could you please correct this.
Shahrooz Faghihroohi sagt:
Hi Fabian, thanks for mentioning that! I forgot to complete this part. The emphasis would be on:
Problem Statement / Big picture
Methodology of other works related to the topics
Important Results that may compare the state-of-the-art related to the topic.