MPG ETH CLS Student Projects


75472308 036a 4f15 81be bf6791839a4b.png 500 500

Improving Feedback in Human-in-the-loop Reinforcement Learning

In this project, we would like to explore whether an attention or saliency visualization technique can improve teacher feedback in a human-in-the-loop reinforcement learning scenario.

Reinforcement Learning Computer Vision Saliency Attention Human-in-the-loop

Read more



A7bd18d2 089c 4cb5 935b fa0aa281ad1b.png 500 500

Machine Learning-based Lake Ice Monitoring using UAVs

This thesis will focus on an inter-disciplinary (intersection of artificial intelligence and climate change) research topic. This work is part of the project “Integrated lake ice monitoring and generation of sustainable, reliable, long time series”.

Lake Ice Detection Few-Shot Learning Deep Learning Random Forests Remote Sensing Computer Vision Image Analysis Climate Change Drone Image Processing UAV

Read more



2a47ec50 3adb 4ba6 8b56 c519de087c0f.jpg 500 500

Automatic detection of fishes in underwater photogrammetric images

In underwater photogrammetric surveys, fishes are often visible in the images and disturb disturbing orthophotos and model textures. The goal of the project is to design and implement an automatic system for detecting fishes and masking them out.

Photogrammetry Computer Vision Object Detection Machine Learning Underwater Archaeology

Read more



0152f4e8 01d8 47dd a0fe 5122aedb5f46.png 500 500

Using machine learning to detect and predict changes of land use from aerial imagery

The ability to understand the evolution of land use (e.g. construction, change of land type) is crucial in fields such as urban planning, agriculture, natural resources management, and even autonomous flight. We will develop a Bayesian Recurrent Neural Network approach.

Bayesian Deep Learning Recurrent Neural Networks Remote Sensing Agriculture Environmental Sciences Computer Vision

Read more



4c46a8b1 9a2a 476e 8a1c b2791484f0a6.png 500 500

Image - Semantics Consistency

In this project we will exploit generative models (e.g. GANs) to provide introspection to semantic segmentation networks in open-world settings as faced by mobile robots.

Deep Learning Generative models Semantic Segmentation Robotics Scene Understanding

Read more