Speaker: Valentin Bickel (MPI Göttingen, ETH)
Website: https://www.mps.mpg.de/person/59397

Talk title: Deep Learning-driven Mapping of Lunar and Martian Mass Wasting Features: Methods and Applications

Missions like the Lunar and Mars Reconnaissance Orbiters returned more than 2 million high-resolution satellite images of the surface of the Moon and Mars. This data archive is too big for manual examination and thus contains a large amount of valuable, yet undiscovered information about the recent lunar and martian geologic evolution. For example, the occurrence and distribution of mass wasting features such as landslides and rockfalls might be indicative for a planet’s or moon’s endo- and exogenic activity. We are applying a state-of-the-art deep learning-driven object detection method to identify and map mass wasting features on planetary scales in the available big data and use the results to better understand the geomorphic activity and evolution of our Moon and the red planet. This seminar talk will cover the applied methodology and will report the first and preliminary results of our work.

Office: G6, 204

Host: Daniel Angerhausen (hc.ebinu.hsc|nesuahregna.leinad#hc.ebinu.hsc|nesuahregna.leinad)


Fill out this table in order to set up a meeting with our CSH Colloquium Speaker.
Please include your name, time, and the room, in which you would like to meet.
You may sign up for two consecutive slots if time allows.
Do not edit anyone else's entry without their permission!

Thursday, Feb 20, 2020
10.00–11.00: Daniel Angerhausen (G6, 015)
11.00–12.00: N. Thomas, P. Becerra (G6, -107)
12.15–13.00: Talk
13.00–14.00: Lunch
14.15–15:30: A. Valantinas (G6, 212)

Dinner plans: Wed 20:00 - Injera (contact DA if you want to join)