We meet every Monday, at 10 AM in MI^2 DataLab (room 044, Faculty of Mathematics and Information Science, Warsaw University of Technology) or online.
Join us at https://meet.drwhy.ai (or https://meet.google.com/nno-okiz-bxy).
2022-12-19
Abstract: In the age of Deep Learning in the field of Computer Vision there still exist several classical algorithms that are perfect for segmentation task, like active contours. When they are perfect for single cases with human supervision (like in photoshop) they struggle when it comes to performance on scale. However, is there a way to find relationship between image features and algorithm's parameters and how do they influence each other? On the next seminar I would like to give presentation on my recent research towards answering this question, where classical algorithms will meet with autoML.
- 2022-10-03 - Organisation of the seminar
- 2022-10-10 - xLungs - Efficient and Explainable Diagnostics’ Support for Radiologists
- 2022-10-17 - HOMER - Achieving Responsible ML by Human Oriented autoMated machinE leaRning
- 2022-10-24 - ATLAS - Automated documenT anaLysis for sociAl awareneSs
- 2022-11-07 - Explainable machine learning for survival analysis - Mateusz Krzyziński, Mikołaj Spytek, Hubert Baniecki
- 2022-11-14 - Lungs vector representations - Maciej Chrabąszcz, Bartek Sobieski, Weronika Hryniewska-Guzik
- 2022-11-21 - ATLAS: Explaining abstractive summarization - Emilia Wiśnios
- 2022-11-28 - How to organize our work - discussion
- 2022-12-05 - Cancelled
- 2022-11-12 - research presentation: HOMER
- 2022-11-19 - research presentation: xLungs
- 2023-01-02 - update: ?HOMER
- 2023-01-09 - update: ?xLungs
- 2023-01-16 - update: ?ATLAS
- 2023-01-23 - Summary