Open Semester Projects and Thesis
Note: Topics labeled as “semester project” are suitable for the “Interdisciplinary Project” course in the Geomatics MSc curriculum 2013, respectively for the “Geomatics Project I” or “Geomatics Project II” in the MSc curriculum 2022.
Beyond Marigold: Diffusion-Based Monocular Predictor
Extend the recent Marigold in different aspects
Keywords
Diffusion model, monocular predictor
Labels
Master Thesis
Description
Goal
Contact Details
More information
Open this project... call_made
Published since: 2024-03-18 , Earliest start: 2024-04-01 , Latest end: 2024-11-30
Organization Computer Vision and Geometry Group
Hosts Peng Songyou , Obukhov Anton , Ke Bingxin
Topics Information, Computing and Communication Sciences
CAD2Real2CAD: Domain Adaption for LiDAR Novel View Synthesis
LiDAR simulation is important for developing and testing autonomous vehicles [3]. Neural LiDAR FIeldsNFL like [1] took a step towards simulating realistic LiDAR scans from novel viewpoints, by synthesizing LiDAR point clouds with the neural field method. However, they are currently limited to static scenes and are not yet able to leverage the large volume of existing 3D assets (like CAD models of vehicles) to improve the diversity of the simulated scenes, e.g., by inserting or deleting individual objects in such a way that their LiDAR reflections are realistic , and compatible with the rest of the scene. The goal of the project is to learn from large volumes of data how to synthesize such realistic LiDAR patterns, to enable highly accurate simulation of LiDAR scans for dynamic scenes.
Keywords
LiDAR simulation, autonomous driving 3D deep learning, neural fields
Labels
Semester Project , Master Thesis , ETH Zurich (ETHZ)
Description
Goal
Contact Details
More information
Open this project... call_made
Published since: 2024-02-12 , Earliest start: 2023-11-20 , Latest end: 2024-11-30
Organization Photogrammetry and Remote Sensing (Prof. Schindler)
Hosts Huang Shengyu
Topics Information, Computing and Communication Sciences
Depth Completion Using Advanced Anisotropic Diffusion Techniques
In this project, we aim to develop a depth completion framework based on the anisotropic diffusion prior.
Keywords
Depth Completion, Deep Learning, Machine Learning, Computer Vision
Labels
Master Thesis
Description
Goal
Contact Details
More information
Open this project... call_made
Published since: 2023-12-05 , Earliest start: 2024-02-26 , Latest end: 2024-08-31
Applications limited to ETH Zurich
Organization Photogrammetry and Remote Sensing (Prof. Schindler)
Hosts Metzger Nando
Topics Information, Computing and Communication Sciences , Engineering and Technology