Skip to the content.

Program

Time Talk Speaker Slides
1:30 - 1:35 Opening Timo Milbich Slides
1:35 - 2:00 Introduction to Similarity and Deep Metric Learning
(20 min + 5 min QA)
Björn Ommer Slides
2:00 - 2:45 DML I: The Family of Objective Functions in DML
(40 min + 5 min QA)
Timo Milbich Slides
2:45 - 3:15 Coffee break    
3:15 - 4:00 DML II: Beyond Pairs and Triplets - Contextual classification losses
(40 min + 5 min QA) (virtual)
Ismail Elezi Slides
4:00 - 4:30 Best Practices in DML - Dataset, Metrics, Tricks, Reality Check
(25 min + 5 min QA)
Jenny Seidenschwarz Slides
4:30 - 4:50 Towards Realistic Evaluation of OOD Generalization in DML
(15 min + 5 min QA)
Timo Milbich Slides
4:50 - 5:10 Connections between DML and other fields
(15 min + 5 min QA)
Jenny Seidenschwarz Slides