ETH Zurich - D-INFK - IVC - CVG - Lectures - Deep Learning Seminar - Course 2020

Course 2020

Course title: Deep Learning for Computer Vision: Seminal Work
Course ID: 263-5904-00L
Lecturers: Dr. Martin Oswald, Dr. Zhaopeng Cui
Teaching assistants: Daniel Thul, Ian Cherabier, Mihai Dusmanu, Marcel Geppert, Sandro Lombardi, Luca Cavalli, Rémi Pautrat, Taein Kwon, Songyou Peng, Denys Rozumnyi, Katarina Tóthová, Silvan Weder, Zuoyue Li, Paul-Edouard Sarlin
Venue: Mo 15-17h in 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.


  • 24.02. Paper assignment has been published.
  • 17.02. Paper assignment will open today at 5pm. Please choose your favorite papers in the Moodle course until the end of Feb. 21 (this Friday).
  • 17.02. The slides of today's introduction session are online now.

Important information

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. Each student will be graded based on both their presentation (80%) and their participation in the assigned discussions (20%). Attendance is required to pass the course (3 absences allowed).


Schedule (Tentative)

Date Topic Presenter Critics Assistant
17.02. Introduction
24.02. No seminar - Paper assignment will be announced on this page
02.03. 01 Learning representations by back-propagating errors Marcel Geppert
02.03. 02 Backpropagation applied to handwritten zip code recognition Adrian K. Costanza Maria I., Frédéric O. Denys Rozumnyi
02.03. 03 ImageNet Classification with Deep Convolutional Neural Networks Boyan D. Carla J., Florin V. Dr. Zhaopeng Cui
09.03. 04 U-Net: Convolutional Networks for Biomedical Image Segmentation Denys Rozumnyi
09.03. 05 Deep residual learning for image recognition Christian B. Simon H., Jiahao W. Mihai Dusmanu
09.03. 06 Xception: Deep Learning with Depthwise Separable Convolutions Valentin W. Matej S., Ali A. Luca Cavalli
16.03. 07 Dropout: a simple way to prevent neural networks from overfitting Erick T. Rishabh S., Yannick R. Daniel Thul
16.03. 08 Group Normalization Frédéric O. Jiahao W., Boyan D. Zuoyue Li
23.03. 09 Visualizing and Understanding Convolutional Networks Yannick R. Valentin W., Erick T. Silvan Weder
23.03. 10 What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision? Yannick S. Rishabh S. Rémi Pautrat
30.03. 11 Human-level control through deep re-inforcement learning Matej S. Erick T., Simon H. Luca Cavalli
30.03. 12 Conditional Random Fields as Recurrent Neural Networks Ian Cherabier
06.04. 13 Learning Convolutional Neural Networks for Graphs Simon H. Frédéric O., Yuqing C. Zuoyue Li
06.04. 14 Dynamic Routing Between Capsules Rishabh S. Boyan D., Matej S. Dr. Martin Oswald
27.04. 15 A Style-Based Generator Architecture for Generative Adversarial Networks Florin V. Dexin Y., Costanza Maria I. Rémi Pautrat
27.04. 16 Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks Yuqing C. Yannick S., Valentin W. Songyou Peng
04.05. 17 Mask R-CNN Costanza Maria I. Yannick R., Adrian K., Carla J. Taein Kwon
04.05. 18 AtlasNet: A Papier-Mâché Approach to Learning 3D Surface Generation Carla J. Yuqing C., Manuel B. Sandro Lombardi
11.05. 19 Working hard to know your neighbor's margins: Local descriptor learning loss Dexin Y. Florin V., Georges P., Ali A. Paul-Edouard Sarlin
11.05. 20 DSAC-differentiable RANSAC for Camera Localization Jiahao W. Manuel B., Christian B., Yannick S. Mihai Dusmanu
18.05. 21 DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation Ali A. Adrian K., Dexin Y., Christian B. Katarina Tóthová
18.05. 22 Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representationsh Sandro Lombardi
25.05. 23 The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks Daniel Thul
25.05. 24 Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation Silvan Weder

© CVG, ETH Zürich webmaster: