Seminar Robot Perception for Navigation - WS 2022/23

Seminar Robot Perception for Navigation

Requirements & Information

• Organizer: Dr. Tim Welschehold
• Co-Organizers: Johannes Meyer, Adriana Gomez, Jannik Zürn, Kürsat Petek, Chenguang Huang Shengshao Yan
• The first meeting will take place on Monday, October 17th 2022 at 15:00-16:00 in an online zoom meeting link
• We had to re-create the seminar in HISinOne due to the problems with the different exam regulations (Pruefungsordnungen). You can find the seminar here (HisInOne-ID: 11LE13S-7307-MB)
• Students are requested to:
  • Write a two-page abstract (prior to the talk) of the assigned paper.
  • Prepare a talk of 20 minutes.
  • Write a finalized three-page abstract, in which they additionally discuss connections from their paper to others presentend in the seminar.
  Everything is supposed to be done in English.
• The two-page summary is due on Wednesday, December 14th, 2022.
• The Seminar will be held as a "Blockseminar" in the end of the Semester. The slides should be discussed with the supervisor two weeks before the presentation. The Presentations will take place on TBA.
• The final-report is due two weeks after the presentation and should be 3 pages long at maximum (latex, a4wide, 11pt) not counting the bibliography and figures. Significantly longer summaries will not be accepted. A LaTeX template can be downloaded here.

---

• Update (2022-10-25):After you have been assigned to the seminar. Please send us your paper priorities.
  1. For each paper 1..n assign the paper a priority 1..n with 1 being the highest priority and n being the lowest.
  2. Use the following format when sending in your priorities: ["YOURNAME", priority_paper_1, priority_paper_2, ..., priority_paper_13]. if you only want to rank 4 papers, just assign priority 13 to all the other papers.
  3. Send E-Mail with your paper priority to robot-perception-navigation@cs.uni-freiburg.de
  4. Example: Your favorite papers are 5, 3, 2, and 8. Please send us ["YOURNAME", 13, 3, 2, 13, 1, 13, 13, 4, 13, 13, 13, 13, 13]

Slides

• Introductional Slides (PDF)
• Giving a Presentation (PDF)

List of Seminar Topics:

1. SegContrast: 3D Point Cloud Feature Representation Learning Through Self-Supervised Segment Discrimination
2. Unsupervised Class-Agnostic Instance Segmentation of 3D LiDAR Data for Autonomous Vehicles
3. Real-Time Multi-Modal Semantic Fusion on Unmanned Aerial Vehicles
4. CMX: Cross-Modal Fusion for RGB-X Semantic Segmentation with Transformers
5. Semantic Segmentation for Thermal Images: A Comparative Survey
6. Lane-Level Street Map Extraction from Aerial Imagery
7. Objects that Sound
8. Learning robust perceptive locomotion for quadrupedal robots in the wild
9. Learning High-Speed Flight in the Wild
10. BEVFusion: A Simple and Robust LiDAR-Camera Fusion Framework
11. Unsupervised Semantic Segmentation by Contrasting Object Mask Proposals
12. Grounded Language-Image Pre-training
13. Scaling Open-Vocabulary Image Segmentation with Image-Level Labels