Final Theses

At our institute, we constantly offer topics for final theses. The topic suggestions you can find on this page are usually strongly related to our assistants' current research, and thus allow you to work on state-of-the-art technologies. Don't hesitate to contact the responsible assistants regarding the topics suggestions or their research in general.

If you have question regarding the organization of final theses, you can contact Dominik Mehlem.

Here is a list of the currently offered topics:

Virtual Human’s Behavior in Social Mixed Reality

Virtual humans are digital representations of people. One of the main challenging aspects is that they act and move in a coherent and believable manner, adhering to the affordances of AR or VR environments. Here, virtual human’s fidelity can be explored in three major aspects, physical (how they look, sound, and move), functional (how they react and interact with the environment, which is particularly important for AR), and psychological (the level of authenticity in the emotions they display). We offer several master's theses that can address one or several of these aspects. The students interested in pursuing these topics should have or be willing to acquire strong skills in 3D design and development (Unity, Unreal, Blender, Mixamo) and should be interested in conducting an empirical study and familiarizing with research methods in human-computer interaction.

Suitable for Bachelor and Master theses. Contact person: Dr.-Ing. Stephanie Gabriela Arévalo Arboleda

Subjective Quality Assessment of AI-based Videos

We offer theses in the field of quality assessment of AI-generated and AI-processed videos. Topics include the design of subjective user studies, the analysis of subjective data, and the development of evaluation models, with the goal of algorithmic prediction of human perception. Students gain practical experience with modern video generation or processing techniques, the design of experimental studies, and statistical data analysis. Projects can be tailored to individual interests and prior knowledge.

Suitable for Bachelor and Master theses. Contact person: Julius Prenzel, M.Sc.

Point Cloud Compression

The general idea of point cloud compression is to find a representation of a point cloud which is less data exhaustive than the original raw point cloud. Currently, in MPEG (Moving Picture Experts Group) there are two different point cloud compression standards being developed, which will most likely be the new state-of-the-art for the next years. One standard is called video-based point cloud compression (V-PCC) and focuses on the compression of dynamic point cloud sequences (i.e. point cloud “video”). Here, the point cloud is projected onto a 2D canvas and the so created video streams are then compressed by a conventional video coder. The other standard is called geometry-based point cloud compression (G-PCC) and currently focuses on compression of static point clouds. Here, the point cloud is compressed with the help of conventional 3D tools, i.e. the use of an octree to iteratively split the point cloud into 8 subblocks.

Machine/Deep learning-based Approaches for Point Cloud Compression	Dominik Mehlem, M.Sc.
In recent years, first approaches for autoencoder-based approaches showed to outperform the traditional approaches from a RD-based perspective. Still these approaches show promising opportunities for further research. Possible topics could include, but are not limited to: Improve on the already proposed autoencoder architectures to more efficiently compress the geometry/texture and/or make them more efficient in terms of computational complexity Test and/or design perceptual cost functions to further improve the training of networks Design architectures and/or approaches to code the geometry as well as the attributes (for example the color information of the point cloud) in a combined manner Provide hybrid approaches with the conventional point cloud compression standards that allow for scalability (in quality and resolution/size) to benefit of the advantages of both worlds Over time there might be different and more diverse topics be available in this area. When interested please feel free to reach out to me, so that we can further discuss details.
Suitable for Bachelor and Master theses.

Machine/Deep learning-based Approaches for Point Cloud Compression

Dominik Mehlem, M.Sc.

Example frames of dynamic point cloud sequences

In recent years, first approaches for autoencoder-based approaches showed to outperform the traditional approaches from a RD-based perspective. Still these approaches show promising opportunities for further research.

Possible topics could include, but are not limited to:

Improve on the already proposed autoencoder architectures to more efficiently compress the geometry/texture and/or make them more efficient in terms of computational complexity
Test and/or design perceptual cost functions to further improve the training of networks
Design architectures and/or approaches to code the geometry as well as the attributes (for example the color information of the point cloud) in a combined manner
Provide hybrid approaches with the conventional point cloud compression standards that allow for scalability (in quality and resolution/size) to benefit of the advantages of both worlds

Over time there might be different and more diverse topics be available in this area. When interested please feel free to reach out to me, so that we can further discuss details.

Suitable for Bachelor and Master theses.