Instructors: Prof. Dr. Nassir Navab, Anees Kazi, Roger Soberanis, Mahsa Ghorbani
Announcements
- 13.12.2021 - The recorded video of the first lecture has been added to the schedule.
- 22.10.2021 - The schedule has been released.
- 13.10.2021 - Let us know your four preferred papers by email by Sun. 17.10
- 13.10.2021 - List of topics has been published.
- 12.10.2021 - The list of papers will be published on October 13th. Further details will be sent to the participants via email.
- 12.10.2021 - The classes will start on November 2nd, 12:00 pm - 2 pm CET.
- 15.07.2021 - Registrations are open from 15.07.2021 to 20.07.2021 through the TUM Matching Platform. Additionally, let us know your interest through our application form.
- 12.07.2021 - The slides of the preliminary meeting are available here (slides).
- 05.07.2021 - Course Wiki is up!
- 05.07.2021 - Contact information. If you have any questions about the seminar, feel free to contact Roger Soberanis (roger.soberanis@tum.de) or Anees Kazi (anees.kazi@tum.de)
Introduction
Graph Deep Learning is a new exotic branch in many fields like computer Vision and Medical Imaging. Many real-world medical and non-medical datasets can be represented in the form of graphs, providing a powerful source of information for machine learning models. This graph-based data, combined with the success of the convolutional neural networks, has motivated to translate the key ingredients of deep learning models into the graph domain. Many communities such as healthcare, social media, and computer vision are moving towards analyzing the data using Graph Convolutions. This seminar provides a space for discussion of the recent scientific publications on GCN with a focus on their medical applications and others.
Registration
- Interested students should attend the introductory meeting on 07.07.2021 to enlist in the course. Details are available at TUM online.
- Students can only register through TUM Matching Platform themselves if the maximum number of participants hasn't been reached (please pay attention to the Deadlines).
- A maximum number of participants: 20.
Requirements and Course Structure
For this course, previous background in Machine/Deep Learning is required.
During the course, students will select and present a scientific publication from the given list. Additionally, students will also contribute to the course blog post with their own summary of the selected article. Requirements are described in the image below:
List of Topics
| Topic | Number | Title | Journal/Conference | Tutor | Student | Link |
|---|---|---|---|---|---|---|
| Registration and reconstruction | 2 | GraphRegNet: Deep Graph Regularisation Networks on Sparse Keypoints for Dense Registration of 3D Lung CTs | TMI 2021 | Farid | Nicolas Peter | https://ieeexplore.ieee.org/abstract/document/9406964 |
| Registration and reconstruction | 3 | Image-to-Graph Convolutional Network for Deformable Shape Reconstruction from a Single Projection Image. | MICCAI 2021 | Mahdi | Stella Dimitra | https://arxiv.org/pdf/2108.12533.pdf |
| Registration and reconstruction | 4 | Pose2Mesh: Graph Convolutional Network for 3D Human Pose and Mesh Recovery from a 2D Human Pose | ECCV 2020 | Lennart | Dominik | https://arxiv.org/pdf/2008.09047.pdf |
| Interpretability | 5 | INTERPRETING GRAPH NEURAL NETWORKS FOR NLP WITH DIFFERENTIABLE EDGE MASKING | ICLR 2021 open review | Yousef | Jingpei | https://openreview.net/forum?id=WznmQa42ZAx |
| Interpretability | 6 | Should Graph Convolution Trust Neighbors? A Simple Causal Inference Method | Sigir 2021 | Ashkan | Armin | https://dl.acm.org/doi/pdf/10.1145/3404835.3462971 |
| Interpretability | 7 | Disentangled Graph Convolutional Networks | International Conference on Machine Learning. PMLR, 2019 | Azade | Ahmed Alaaeldin Fathy Hanafy | http://proceedings.mlr.press/v97/ma19a/ma19a.pdf |
| General | 8 | Multi-label zero-shot learning with graph convolutional networks | Neural Networks 2020 | Alaa | Mohamed Taieb | https://www.sciencedirect.com/science/article/abs/pii/S0893608020303336 |
| General | 9 | Learning Graph Convolutional Networks for Multi-Label Recognition and Applications | TPAMI 2021 | Mahsa | Simay | https://ieeexplore.ieee.org/abstract/document/9369105 |
| Medical - Brain | 10 | BrainGNN: Interpretable Brain Graph Neural Network for fMRI Analysis | MedIA 2021 | Shahrooz | Amine | https://www.biorxiv.org/content/10.1101/2020.05.16.100057v4.full.pdf |
| Medical - Brain | 11 | Attention-Guided Deep Graph Neural Network for Longitudinal Alzheimer’s Disease Analysis | MICCAI 2020 | Mathias | Sergei | https://link.springer.com/chapter/10.1007/978-3-030-59728-3_38 |
| Medical - Brain | 12 | A mutual multi-scale triplet graph convolutional network for classification of brain disorders using functional or structural connectivity | TMI 2021 | Shahrooz | Nicolas Robert Baptiste | https://ieeexplore.ieee.org/abstract/document/9324801 |
| Medical - Brain | 13 | Multi-Head GAGNN: A Multi-Head Guided Attention Graph Neural Network for Modeling Spatio-Temporal Patterns of Holistic Brain Functional Networks | MICCAI 2021 | Anees | Erekle | https://link.springer.com/chapter/10.1007%2F978-3-030-87234-2_53 |
| Medical - Segmentation and Classification | 14 | SGNET: Structure-Aware Graph-Based Networks for Airway Semantic Segmentation | MICCAI 2021 | Roger | Janik | https://link.springer.com/chapter/10.1007%2F978-3-030-87193-2_15 |
| Medical - Segmentation and Classification | 15 | Hybrid graph convolutional neural networks for landmark-based anatomical segmentation | MICCAI 2021 | Tariq | Farid | https://arxiv.org/abs/2106.09832 |
| Medical - Segmentation and Classification | 16 | Early Detection of Liver Fibrosis Using Graph Convolutional Networks | MICCAI 2021 | Mahsa | Sener | https://link.springer.com/chapter/10.1007/978-3-030-87237-3_21 |
Schedule
Sessions will take place Tuesdays from 12:00 to 14:00 hrs. A group of Two to Three students will present their papers during each session.
The schedule will be released on 19.10. Sessions will start on 02.11 with introductory lectures by the organizers.
Place will be confirmed during the week of 25 of October.
| Date | Time | Place | Topic | Tutor (email) | Student | Additional Info/ slides |
|---|---|---|---|---|---|---|
| 2.11.21 | 12:00 - 14:00 | Intro to GDLMA I | Video Lecture | |||
| 9.11.21 | 12:00 - 14:00 | Intro to GDLMA II | ||||
16.11.21 | 12:00 - 14:00 | GraphRegNet: Deep Graph Regularisation Networks on Sparse Keypoints for Dense Registration of 3D Lung CTs | Farid (mf.azampour@gmail.com) | Nicolas Peter | ||
| Image-to-Graph Convolutional Network for Deformable Shape Reconstruction from a Single Projection Image. | Mahdi (m.saleh@tum.de) | Stella Dimitra | ||||
| Pose2Mesh: Graph Convolutional Network for 3D Human Pose and Mesh Recovery from a 2D Human Pose | Lennart (lennart.bastian@tum.de) | Dominik | ||||
23.11.21 | 12:00 - 14:00 | |||||
| Project presentation by CAMP: Graph Deep Learning for Healthcare Applications. | Dr. Anees Kazi | |||||
30.11.21 | 12:00 - 14:00 | INTERPRETING GRAPH NEURAL NETWORKS FOR NLP WITH DIFFERENTIABLE EDGE MASKING | Yousef (ashkan.khakzar@tum.de ) | Jingpei | ||
| Should Graph Convolution Trust Neighbors? A Simple Causal Inference Method | Ashkan (ashkan.khakzar@tum.de ) | Armin | ||||
| Disentangled Graph Convolutional Networks | Azade (azade.farshad@tum.de ) | Ahmed Alaaeldin Fathy Hanafy | ||||
7.12.21 | 12:00 - 14:00 | |||||
| Project presentation by CAMP | Dr. Hendrik Burwinkel | |||||
14.12.21 | 12:00 - 14:00 | Multi-label zero-shot learning with graph convolutional networks | Alaa / Anees (anees.kazi@tum.de) | Mohamed Taieb | ||
| Learning Graph Convolutional Networks for Multi-Label Recognition and Applications | Mahsa (mahsa.ghorbani@tum.de) | Simay | ||||
| BrainGNN: Interpretable Brain Graph Neural Network for fMRI Analysis | Shahrooz (hahrooz.faghihroohi@tum.de) | Amine | ||||
21.12.21 | 12:00 - 14:00 | |||||
| Project presentation by CAMP | ||||||
28.12.21 | 12:00 - 14:00 | |||||
| No Session | ||||||
4.01.22 | 12:00 - 14:00 | |||||
| No Session | ||||||
11.01.22 | 12:00 - 14:00 | Attention-Guided Deep Graph Neural Network for Longitudinal Alzheimer’s Disease Analysis | Mathias (matthias.keicher@tum.de) | Sergei | ||
| A mutual multi-scale triplet graph convolutional network for classification of brain disorders using functional or structural connectivity | Shahrooz (hahrooz.faghihroohi|[at]|tum.de) | Nicolas Robert Baptiste | ||||
| Multi-Head GAGNN: A Multi-Head Guided Attention Graph Neural Network for Modeling Spatio-Temporal Patterns of Holistic Brain Functional Networks | Anees (anees.kazi@tum.de) | Erekle | ||||
18.01.22 | 12:00 - 14:00 | |||||
| Project presentation by CAMP | ||||||
25.01.22 | 12:00 - 14:00 | SGNET: Structure-Aware Graph-Based Networks for Airway Semantic Segmentation | Roger (roger.soberanis@tum.de) | Janik | ||
| Hybrid graph convolutional neural networks for landmark-based anatomical segmentation | Tariq (t.bdair@tum.de) | Farid | ||||
| Early Detection of Liver Fibrosis Using Graph Convolutional Networks | Mahsa (mahsa.ghorbani@tum.de) | Sener | ||||
1.02.21 | 12:00 - 14:00 | |||||
| Project presentation by CAMP | ||||||
Resources and Material
An introductory lecture to GCNs by Prof. Xavier Bresson on Alfredo Canziani Youtube channel:
If you find more interesting material you would like to share in this section, feel free to contact us.