Author: Frederike Dümbgen

TAKEN: Experimental evaluation of unsupervised indoor localization

Past Projects

Contact: Dümbgen, Frederike, Sepand Kashani Synopsis: Experimental evaluation of unsupervised indoor localization. Level: Summer internship or BS/MS project Description: Due to their aptitude in depicting complex dependencies, neural networks are a promising candidate for indoor localization. Omnipresent phenomena such as multi-path signal propagation, shadowing and device noise introduce high noise in distances inferred from radio (…)

Open-source hardware and GUI design for a blind drone

Past Projects

Contact: Dümbgen, Frederike Synopsis: Design a PCB and user interface for a Crazyflie drone to process acoustic and RF signals. Level: BS, MS Description: The Crazyflie (https://www.bitcraze.io/crazyflie-2-1/ ) is a developer-friendly, open-source drone solution. It comes with a customizable python client (including GUI), and open-source firmware which is readily extendable. In this semester project, we (…)

Learning acoustics-based localization of a blind drone

Past Projects

Contact: Dümbgen, Frederike Synopsis: Study and test drone’s capability to localize itself from acoustic signals. Level: BS, MS Description: Acoustic signals are inexpensive to create and record and their physics are well understood. Using only relatively cheap and light hardware, drones could therefore be equipped with audio processing capabilities. However, very few drone localization systems (…)

Angle-based drone swarm configuration recovery

Available Projects

Contact: Dümbgen, Frederike Synopsis: Use off-the-shelve Bluetooth angle of arrival (AoA) technology to recover the relative positions of drones in a swarm. Level: BS, MS Description: Swarms of multiple cheap and light-weight drones have become a great alternative to expensive heavy-duty UAVs. Drone swarms can reconfigure on the fly to perform complex tasks requiring for (…)

Learning multi-modal localization

Past Projects

Contact: Dümbgen, Frederike Synopsis: Development of a learning-based framework for indoor localization using different modalities. Level: MS Description: Due to their aptitude in depicting complex dependencies, neural networks are a promising candidate for indoor localization. Omnipresent phenomena such as multi-path signal propagation, shadowing and device noise introduce non-linear effects in the data, and make conventional (…)

Bring voice user interfaces to our offices

Past Projects

Contact: Dümbgen, Frederike ; Hoffet, Adrien Synopsis: Implement a new way of interacting with your computer via voice control instead of the mouse and keyboard. Level:BS, MS Description: Google Home and Amazon Alexa are quickly revolutionizing how we interact with smart devices. Both use “wake words” (“OK Google” and “Alexa” respectively) to detect the user’s (…)