ETH Zurich - D-INFK - IVC - CVG - Lectures - Deep Learning Seminar

Deep Learning Seminar


Course title: Deep Learning for Computer Vision: Seminal Work
Course ID: 263-5904-00L
Lecturers: Dr. Iro Armeni
Teaching assistants: Daniel Thul, and Luca Cavalli, Mihai Dusmanu, Jonas Hein, Taein Kwon, Zuoyue Li, Denys Rozumnyi, Rémi Pautrat, Songyou Peng Sandro Lombardi, Paul-Edouard Sarlin, Silvan Weder
Venue: Mo 16-18h, CAB G 57

This seminar covers seminal papers on the topic of deep learning for computer vision. The students will present and discuss the papers and gain an understanding of the most influential research in this area - both past and present. The objectives of this seminar are two-fold. Firstly, the aim is to provide a solid understanding of key contributions to the field of deep learning for vision (including a historical perspective as well as recent work). Secondly, the students will learn to critically read and analyse original research papers and judge their impact, as well as how to give a scientific presentation and lead a discussion on their topic.

Each student chooses one paper from the provided collection to present during the course of the seminar. The students will be supported in the preparation of their presentation by the seminar assistants.

Important information

  • Class Material: Class material will be posted on Moodle. We will also use Moodle to submit assignments.
  • Assignments: Each student will present one paper. Everybody is encouraged to read each paper before it is being presented and engage in a discussion following the presentations. To foster interesting discussions, each paper will also be assigned two "critics" who study the paper and prepare questions for the discussion.
  • Grading: Each student will be graded based on both their presentation (75%) and their participation in the assigned discussions (20%). There is a small participation grade (5%) for those that ask questions in papers even if they are not assigned to them.
  • Attendance: Attendance is required to pass the course (3 absences allowed).
  • The class is held in person, except if otherwise stated.

Schedule

Date Topic Presenter Critics Assistant
21.02. Introduction
28.02. #01. ImageNet Classification with Deep Convolutional Neural Networks Adam K. Markus P., Jiasong G. Jonas Hein
28.02. #02. Visualizing and Understanding Convolutional Networks Rares C. Alexandra S., Yitong X. Mihai Dusmanu
07.03. #03. U-Net: Convolutional Networks for Biomedical Image Segmentation Markus P. Qi C., Kaishuo Z. Denys Rozumnyi
07.03. #04. Deep residual learning for image recognition Jiasong G. Adam K., Rares C. Denys Rozumnyi
14.03. #05. MobileNetV2: Inverted Residuals and Linear Bottlenecks Zeren J. Xiyi C., Markus P. Daniel Thul
14.03. #06. Attention Is All You Need Gauthier B. Arka M., Alexandra S. Paul-Edouard Sarlin
21.03. #07. KPConv: Flexible and Deformable Convolution for Point Clouds Dimitar B. Aleksandar M., Hande H. Luca Cavalli
21.03. #08. Dynamic Routing Between Capsules Qi C. Rares C., Adam K. Luca Cavalli
28.03. #09. Dropout: a simple way to prevent neural networks from overfitting Alexandra S. Jiasong G. Taein Kwon
28.03. #10. A Metric Learning Reality Check Yuxuan W. Javier S., Aleksandar M. Taein Kwon
04.04. #11. MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels Tomasz Z. Dimitar B., Maurits R.
04.04. #12. Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation Hande H. Gauthier B., Xiyi C. Iro Armeni
11.04. #13. FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search Iro Armeni
11.04. #14. Conditional Random Fields as Recurrent Neural Networks Xiyi C. Dimitar B., Alan P. Zuoyue Li
18.04. No Class
25.04. No Class
02.05. #15. Learning Convolutional Neural Networks for Graphs Arka M. Qi C., Chuqiao L., Tomasz Z. Silvan Weder
02.05. #16. A Style-Based Generator Architecture for Generative Adversarial Networks Javier S. Gauthier B., Zeren J. Rémi Pautrat
09.05. #17. Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks Maurits R. Yuxuan W., Chuqiao L. Zuoyue Li
09.05. #18. DSAC-differentiable RANSAC for Camera Localization Kaishuo Z. Daoji H., Tomasz Z. Rémi Pautrat
16.05. #19. SuperPoint: Self-Supervised Interest Point Detection and Description Alan P. Maurits R., Jiahong W. Paul-Edouard Sarlin
16.05. #20. DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation Chuqiao L. Yitong X., Daoji H. Sandro Lombardi
23.05. #21. AtlasNet: A Papier-Mâché Approach to Learning 3D Surface Generation Aleksandar M. Jiahong W., Kaishuo Z. Sandro Lombardi
23.05. #22. NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis Yitong X. Zeren J., Arka M. Daniel Thul
30.05. #23. Convolutional Occupancy Networks Daoji H. Hande H., Yuxuan W. Songyou Peng
30.05. #24. CaSPR: Learning Canonical Spatiotemporal Point Cloud Representations Jiahong W. Alan P., Javier S. Silvan Weder

© CVG, ETH Zürich webmaster: psarlin@inf.ethz.ch