3D Computer Vision

Dear Students, 

as mentioned, there was an error in the spreadsheet that significantly affected the grades: the weighting for the bonus grades was wrong and did not give 60% weight to the exam but rather 20%. That is why some of you may have been confused that the bonus grade was significantly different than the normal grade. 

We have updated this and double checked everything. Unfortunately, this means that some of your grades have changed for the better and for the worse. Please find the correct version in the spreadsheet. I am sorry for the confusion. 

Here is the grading policy explained: 

As said in the beginning of the lecture, the grades from the parts are accumulated by:
- 20% Assignments 
- 20% Project 
- 60% Exam 

The overall percentage is computed with taking the bonus into account and:
- 40% corresponds to 4.0, less than 40% is failure 
- 85% corresponds to 1.0

Best, 

 

Eddy

 

Dear Students, 

unfortunately, there is one more bug in the spreadsheet that will change the final grades. This will be fixed tomorrow morning. Unfortunately, it means that the final grades will change. 

I need to apologize for this. The examination office did initially not inform us that the grades had to be released before the re-exam and therefore we were in a bit of rush. 

Best, 

 

Eddy

Dear Students, 

the final grades are released in: 
https://docs.google.com/spreadsheets/d/1suDBViLj8DO2Wm_DlKwEnYgBSlVGgx-VU-307gBT-gc/edit?usp=sharing

Please check once more that all your assignment, exam and project grades are correct. 

Grades will be uploaded tomorrow morning, and then you register for the re-exam. 

Best, 

Eddy 

Dear Students, 

some of you let me know some changes in group assignments in the assignments and project. For those that were said to contribute only little to the project, we will give 25% of the credit for the project instead of 100%. 

Important: please check all your group assignments in this spreadsheet until 12pm: 
https://docs.google.com/spreadsheets/d/1suDBViLj8DO2Wm_DlKwEnYgBSlVGgx-VU-307gBT-gc/edit?usp=sharing
(under Tab Final Grades)

After that no more changes will be possible. 

Best, 

 

Eddy

Dear Students, 

the examination office has let me know that your grades have to be finalized before you can register for the re-exam. 

For this reason, the exam insight will be tomorrow from 12pm to 3pm. We will also finalize your project grades, such that your final grade will be in the system Wednesday morning. It is important to note that after that it is impossible to change it on your transcript. 

For registration of the re-exam, you will be able to do so Wednesday or Thursday. 

Best, 

 

Eddy

Dear Students, 

please find information about the project presentations here:
https://docs.google.com/spreadsheets/d/1imQEpU-laCKUuoxVrZT477OLLg9CjenK2lGD-tdpbGI/edit?usp=sharing
in sheet "Project Presentations" 

Important: please check the time for your team and let me know if there are conflicts by emailing cvmp-3dvision@cs.uni-saarland.de

This is just a reminder that the submission deadline is Monday 8am. Please include in your submission: 
- Your source code 
- A short PDF documentation of the formulas and method 
- The output of your method (poses and points) on the test datasets 

Good luck! 

Eddy 

Dear Students, 

if you plan to register for the re-exam, please do so Monday afternoon. The registration is not open yet and will open Monday morning. 

For those of you who just like to repeat the exam to improve your grade, you may do so by registering for the re-exam. Grades will be announced Monday morning as well. 

Best, 

 

Eddy 

Dear Students, 

please note the following: 

• In stage 1 there was a camera pose that was wrong. We have uploaded a new version - please switch to verison 3. 
• We have uploaded the data for stage 2. As I mentioned a couple of times, it was possible to start on stage 2 already with the data from stage 1. The difference is that in stage 2 you are not provided gt correspondences. 
• We will release stage 3 and submission instructions shortly. 

 

Best, 

 

Eddy

Dear Students, 

the exam time and place is:
28th February 2024,  2:00 pm - 5:00 pm
Geb. E2 2, Günter-Hotz-Hörsaal

In general you can find exam times at the bottom of this page:
https://saarland-informatics-campus.de/studium-studies/#study-calendar

Best, 

 

Eddy 

Dear Students, 

Michael Hahn is a close collaborator of our lab and looking for Hiwi to support an exciting project in the intersection of language, vision and neuroscience. The contract can be between 4 and 16 hours per week. Please find the lab website here: 
https://lacoco-lab.github.io/home/

To apply, please send transcripts and CV to:
mhahn@lst.uni-saarland.de

Best, 

 

Eddy

Dear Students, 

we encourage you to come to the office hours to discuss the project, which many have already done. Please note:

• On the upcoming Thursday I will be absent and only our PhD student Raza will be there 
• Next week the office hours will be Monday 4-6pm and Friday 4-6pm instead of Tuesday and Thursday 
• The remaining office hours will Tuesdays and Thursdays 4-6pm 

Best, 

 

Eddy

Dear Students, 

please find the list of admitted students below. 

The following groups have been admitted to the exam: 
30, 19, 26, 7023705, 10, 13, 27, 16, 1, 12, 28, 25, 21, 14, 24, 11, 31, 7, 17

The following groups have not been admitted to the exam:
22

For the groups 13 and 14 we have detected a case of plagiarism. Please come talk to me during the project office hours. 

The following student IDs are admitted to the exam (please check): 
7059080
2551271
2576064
7058726
7059404
7012375
7015701
2581575
7023705
7046952
7026841
7040419
7039059
2578983
7047365
7007584
7028704
7041511
7038998
7043719
7047554
7044248
7039954
7044113
7002316
7046909
7043821
7040005
7024015
7047530
7047619
7047281
7028849
7047804
7047931
7039480
7047939
7023892
7047856
7055226
7012787
7059212
7058116
7056894
7055706
7056998
7047009
7058536
7058343
7047533
7055462
7058867
7058783
7047447

Best,

 

Eddy















 

 

Dear Students, 

please note the following important information: 

• For the stage 1 data there is a scale ambiguity in the evaluation. We will release updated data tomorrow. 
• Stage 2 data will follow shortly. 
• The assignment results will be released on Tuesday during the day. This will determine if you are admitted to the exam. From what we can currently tell, everyone who handed in assignments is likely to be admitted. In the unlikely case you are not admitted, please schedule a personal meeting with me on Tuesday if you would like to discuss the results. 
• For everyone else, the assignment discussion will take place on Friday the 23rd from 1 - 6pm in my office (E17 1.06). 
• Please note that your project grades will not be available for the registration of the re-exam. As discussed with the examination office, everyone in doubt of passing the course should register for the re-exam. When your project grades come out and indicate that you have passed, you will be automatically deregistered from the re-exam. 
• Please note that you need to submit the code of your projects before the presentation on March 18th at 8m. Your submission needs to contain a short write-up explaining the equations in your code. The submission is open in the CMS system. 

Best, 

 

Eddy 

Dear Students, 

it seems there was an issue with the COLMAP assignment. In the virtual machine, you need to execute "conda activate base" instead of "conda activate colmap". 

To this end, we have granted an extension until Monday 8am for the assignment. 

Best, 

 

Eddy 

 

Dear Students, 

as mentioned in the lecture, due to a university event that I am required to attend on Friday, there will be no lecture. 

I have presented the project in the lecture today (see slides and recordings) and the data will be released today as well. The next meeting is the project office hours on Feb 13th from 4-6pm. 

 

Best, 

 

Eddy 
 

Dear Students, 

it has been great to work with all of you so far, and I would like to thank everyone for attending the first 3D computer vision lectures! While it has been the first time for the course, and we know that some content still has to be improved, we would appreciate any feedback from you that helps us improve the course even further in the future. Please take a few minutes to fill out the evaluation form: 

Lecture: 
https://qualis.uni-saarland.de/eva/?l=147024&p=9wbj4h

Tutorial:
https://qualis.uni-saarland.de/eva/?l=1470241&p=6d1o8w

Best, 

 

Eddy

Dear students, 

a few notes regarding the assignment: 
- There is a spurious argument "i" to the function sgm(). Please remove it, it was from some old testing code. 
- You do not need to normalize image values (keep them between 0..255) 
- There was accidentally an "argmin" instead of a "min" in the slides. I will provide an updated version in a few minutes. 

Best, 

 

Eddy 

Dear Students, 

please note that there will be a lecture on the upcoming Friday, 12/22. There will be video recordings available. 

The next exercise will be release on this Friday. You will have the first two weeks to complete it. 

Please note that in the first week of January there are no lectures. 

Best, 

 

Eddy 

Dear Students, 

our student Kunwar Maheep Singh is looking for a group. If there is a remaining spot in any group, please message him at kusi00001@stud.uni-saarland.de and update the Google spreadsheet at: 
https://docs.google.com/spreadsheets/d/1imQEpU-laCKUuoxVrZT477OLLg9CjenK2lGD-tdpbGI/edit?usp=sharing

Best, 

 

Eddy

Dear Students, 

unfortunately, there is a small error in assignment 2. Please note the following correction: 

In Q2, where it says:
"# Rotate the camera such that its z-axis points in the direction of the world x-axis"

this should be the "world y-axis". 

Best, 

 

Eddy

Dear Students, 

tomorrow we will discuss exercise 1 and continue with the lecture. 

Please note that for the exercise discussion, there will be *no* video recordings, as we do not want that people from future semesters have access to the sample solutions. 

Best, 

 

Eddy 

Dear Students, 

thank you for pointing out that there was an issue with the keypoints for tasks 2b and 2c. Please find a zip file with updated keypoints in the materials section and use it accordingly. 

Due to my Covid infection and the update in keypoints, the deadline for the exercise was extended until Monday Dec 4th 8:00am. 

Best, 

 

Eddy

Dear Students, 

unfortunately, I am still struggling with the Covid infection, and there will be no class meeting tomorrow. 

The exercise deadline will be extended until Monday for which I will follow up with a post shortly. 

Monday will then be the exercise discussion and the two missed lectures from last Monday and next Monday will be made available by video. 

I apologize for the inconvenience. 

Best, 

 

Eddy 

 

Dear students, 

due to CVPR deadline which affects myself as well as some of you, there will be a change in the upcoming lectures (please see the calendar):
- Tomorrow (11/13) there will be no lecture. I will be there to answer questions for the exercise if you have any. 
- Next Monday (11/20) there will be lecture 3. 
- Next Friday (11/24) there will be lecture 4. 

Apologies for the delay in releasing the excercise, it will be available soon. 

Best,  

 

Eddy

Dear Students, 

we are still in the process of preparing exercise 1 and will release it tomorrow. 

Upon request, since there is CVPR deadline in a few days, the submission deadline of the exercise will be extended to Dec 1st. I will provide more information in the lecture on Monday. 

Best, 

 

Eddy 

Dear Students, 

this is just a reminder to please register for groups in this spreadsheet:
https://docs.google.com/spreadsheets/d/1imQEpU-laCKUuoxVrZT477OLLg9CjenK2lGD-tdpbGI/edit?usp=sharing

and a reminder to please try out if you can use the virtual machine:
https://drive.google.com/file/d/1Lk4kTEZUhdtf-GBbQxzNxQs1Oo_0fFEU/view?usp=sharing
(password 3dcv_2324)

Best, 

 

Eddy 

Hi everyone, 

the first introductory lecture will take place this Friday at 8:30am in Gebäude E1 3 - Hörsaal III (0.03.1). This lecture will give an overview of the field and the course material. 

After this, there will be a lecture every Monday at 8:30am and a tutorial every Friday at 8:30am in the same location. 

Looking forward to seeing you, 
 

Eddy

Dear Students, 
 

in the end of the lecture, there will be an exciting project on building your own SLAM pipeline. Due to the additional work, we decided to extend the lecture to 9 credit points similar to the lecture Neural Networks: Theory and Implementation. 

Best, 

 

Eddy   

The CVMP lab will offer the new lectures “3D Computer Vision” and “3D Real World Modeling and Inference” in upcoming winter and summer semesters.

The 3D Computer Vision lecture will teach you the fundamentals of going from 2D images to 3D. After having given an intuition of the field and traditional methods, it will proceed to state-of-the-art approaches and provide you with a basis for working in 3D vision. The lecture will cover the mathematical fundamentals, feature matching, stereo, optical flow and scene flow estimation, rigid 3D reconstruction, non-rigid 3D reconstruction and SLAM. 

The 3D Real World Modeling and Inference lecture will teach you how to design models in 3D. It will start with introducing the basic 3D representations ranging from point clouds to meshes, triplanes and voxel grids. Basic and modern deep learning techniques to train and perform inference on these representations will be introduced. Afterwards, the recent implicit representations such as NeRF and DeepSDF with MLPs and how to encode 3D scenes with them efficiently will be covered. Many modern models involve generative models and subsequently the basics of generative models from autodecoders to GANs and finally diffusion models will be presented.