Winter Semester 2025/26, TU München

Organisers: Cecilia Curreli, Mariia Gladkova

Please direct all questions to 4dvision-ws25@vision.in.tum.de.

In this seminar, we will delve into the forefront of computer vision, focusing on the latest deep learning methods for the generation, reconstruction, and prediction of dynamic scenes. As the field progresses, understanding and interpreting dynamic environments have become critical for numerous applications, including autonomous driving, robotics, and augmented reality.

The analyzed recent research papers will highlight significant advancements in dynamic vision. Key areas of focus include: generation of dynamic scenes, reconstruction of dynamic environments, prediction of future states.

The seminar will take place on-site as a block course on the 8th and 9th of January 2026, where each student will present a paper assigned after matching. Attendance is mandatory. Further details can be found in TUMOnline.

The preliminary meeting will take place on Zoom, on the 15th July at 2 pm. The slides of the kick-off meeting will be provided here afterward.

Assignment to the seminar is done via the matching system. In addition to applying to the matching system, send your application documents (transcripts and CV) by the 22nd July (we provide a template email in the preliminary meeting slides). Attending a kick-off meeting is not mandatory for admission.