We are presenting a first and still preliminary schedule of the MIDOG challenge workshop. Congratulation to all accepted presenters! Please Note that all times are UTC! Links to the

Block 1: Introduction and Reference Approach, 14:00-15:20 UTC (16:00 – 17:20 CEST)

14:00-14:20Welcome address and IntroductionMarc Aubreville
14:20-15:05Keynote: How to build trustworthy AI solutions

With the rise of Deep Learning technology and free availability of the required toolchains, developing machine learning (AI) algorithms has become more accessible than ever. In addition, performance of these algorithms is getting better and better, often straight out of the box. Is this sufficient to trust these AI systems enough for deployment in critical application areas as healthcare? What needs to be done to move from an academic proof-of-concept to a validated AI system?

Tobias Heimann
15:05-15:20Domain Adversarial RetinaNet as a Reference Algorithm for the MItosis DOmain Generalization (MIDOG) ChallengeFrauke Wilm

Block 2: Oral Session 1: Cascaded and Multi-stage approaches for mitosis detection
15:30-16:05 UTC (17:30 – 18:05 CEST)

15:30-15:35Cascade RCNN for MIDOG ChallengeSalar Razavi
15:35-15:40Domain Adaptive Cascade R-CNN for Mitosis DOmain Generalization (MIDOG) ChallengeYing Cheng
15:40-15:45Detecting Mitosis against Domain Shift using a Fused Detector and Deep Ensemble Classification Model for MIDOG ChallengeYubo Wang
15:45-15:50Two-step Domain Adaptation for Mitosis Cell Detection in Histopathology ImagesRamin Nateghi
15:50-16:05Q&A and Discussionall

Block 3: Oral Session 2: Instance segmentation-based and adversarial approaches
16:05-16:40 UTC (18:05 – 18:40 CEST)

16:05-16:10Multi-source Domain Adaptation Using Gradient Reversal Layer for Mitotic Cell DetectionSatoshi Kondo
16:10-16:15Sk-Unet Model with Fourier Domain for Mitosis DetectionSen Yang
16:15-16:20Robust Mitosis Detection Using a Cascade Mask-RCNN Approach With Domain-Specific Residual Cycle-GAN Data AugmentationSaima Ben Hadj
16:20-16:25Stain-Robust Mitotic Figure Detection for the Mitosis Domain Generalization ChallengeMostafa Jahanifar
16:25-16:40Q&A and Discussionall

Block 4: Oral Session 3: Augmentation strategies for domain invariance
16:40-17:15 UTC (18:40 – 19:15 CEST)

16:40-16:45Rotation Invariance and Extensive Data Augmentation: a strategy for the Mitosis Domain Generalization (MIDOG) ChallengeMaxime Lafarge
16:45-16:50Assessing domain adaptation techniques for mitosis detection in multi-scanner breast cancer histopathology imagesJack Breen
16:50-16:55Domain-Robust Mitotic Figure Detection with StyleGANYoujin Chung
16:55-17:00MitoDet: Simple and robust mitosis detectionJakob Dexl
17:00-17:15Q&A and Discussionall

Block 5: Results and Awards, Panel Discussion
17:25-18:00 UTC (19:25-20:00 CEST)

17:25-17:40Results and AwardsMarc Aubreville, Katharina Breininger
17:40-18:00Panel DiscussionThe MIDOG Organizers