Betreute Abschlussarbeiten
- Issue Specific Audio Event detection using Deep Learning (Master, October 2019)
- Videobasiertes Situationsbewusstseinssystem mit Datenschutz (Master, July 2020)
- Entwurf und Implementierung eines auf Schwingungssignalen basierenden Kontextwahrnehmungssystems (Master, November 2020)
- Kontextbewusstes Internet der Dinge Plattformdesign und Implementierung (Bachelor, November 2020)
- Schätzung der binären Laufrichtung und der Anzahl der gehenden Personen mithilfe von maschinellem Lernen (Bachelor, November 2020)
- Design und Implementierung eines KI-orientierten 3D Smart Home Simulators (Bachelor, März 2021)
- AI-basierter Streaming-Massenmessungsansatz Entwurf und Implementierung von Schlüsselalgorithmen (Master, TBD)
- DeepWolf: Data set construction toward artificial intelligence-based identity recognition of werewolf (Bachelor, TBD)
- DeepWolf: Artificial intelligence-based identity recognition in the Game of Werewolf (Semantic Analysis Direction) (Master, TBD)
- DeepWolf: Artificial intelligence-based identity recognition in the Game of Werewolf (Intonation, Sentiment Analysis Direction) (Master, TBD)
Topics for Theses
In our research, we are concerned with the AI-based context recognition and context based machine reasoning.
Many different types of sensors provide a wealth of data. This provides the feasibility for the machine to recognize this environment. Traditional situational awareness methods are often based on the setting of rules. It is not easy to perceive more abstract or obscure information. With the efficient implementation of neural network back-propagation algorithms on GPUs, artificial intelligence technology based on deep neural networks has been pushed to the hot spot of the times. Our project uses deep neural networks to learn multi-mode sensor signal data so that the machine can perceive the complex situation in the area where the sensor is located thus offer humans more intelligent supporting services.
With the help of sensors, a digital world can be established that corresponds to the real world. That is the digital twins. In the digital world, computer hardware, computing algorithms, and artificial intelligence methods can be used. Thus, some kind of approach based on this data makes it possible to reason about the real-world situation and requirements. So as to provide better services for mankind in the real-world. Based on the above methods, practical applications such as judicial forensics and smart homes become feasible to benefit humanity.
If you are interested in any of these problems, write me a mail and we can think of a topic that fits into my research area and your skills/interests.
Open Topics
Research Interests
- Machine Learning, Deep Learning, Reinforcement Learning and Machine Learning based Application
- Context-aware Computing
- AI-based Context Recognition
- Cyber-Physical AI Systems
- Pervasive Computing and Internet of Things
Selected Projects
- Research project “Privacy Mechanisms”, funded by Evonik
Academic Services
Journal Reviewer
- IEEE Access
Publikationen
2021 (1)
-
IoT and Deep Learning Based Approach for Rapid Screening and Face Mask Detection for Infection Spread Control of COVID-19In: Applied SciencesDOI: 10.3390/app11083495
2020 (2)
-
Deep Learning-based Vibration Signal Personnel Positioning SystemIn: IEEE Access
-
A Privacy-Protecting Indoor Emergency Monitoring System based on Floor VibrationIn: Adjunct Proceedings of the 2020 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2020 ACM International Symposium on Wearable Computers
2013 (1)
-
Temperature-controlled water saving valveIn: Patent {CN}202812427U
2012 (2)
-
Large powder material storageIn: Patent {CN}202509831U
-
无线局域网络信道规划及控制技术分析In: 中国新通信