Inertial sensors including accelerometers and gyroscopes are key components of nearly each device today, from heavy machinery in industrial settings to personal devices, such as tables, phones, and watches. These sensors and the insights they reveal power much of today’s interactive functions. Examples include activity detection, interaction with the device through gestures and touch, context sensing and more, all of which find application in sports, health, Internet of Things, personal computing, Augmented and Virtual Reality, and others.
At the Sensing, Interaction & Perception Lab (Prof. Christian Holz), we are looking for a PhD candidate with an interest in conducting cutting edge research, with a strong motivation to work on challenging topics, and a strong desire to learn. We will develop novel machine learning-based methods for processing multimodal signals from sensing systems that heavily rely on inertial sensors. We will also explore plenty of applications of such sensor systems and will create interactive prototypes based on embedded technology for mobile and wearable use-cases.
In our group, we have comprehensive experience in methods for multimodal sensing systems and in building embedded systems. We are now looking for a strong candidate to complement our expanding efforts in this space.
Examples of our previous work that leveraged inertial sensors
Virtual Reality/Augmented Reality:
Healthcare and physiological sensing:
The aim of this project is to develop intelligent software experiences that are context sensitive and inform the interactive behavior of user-facing applications on mobile and wearable devices. We will develop novel signal processing methods and user-facing applications for interactive devices with an embedded focus. Our developments will produce novel datasets, computational methods, and the results of in-situ evaluations in interactive scenarios with participants.
Most importantly, candidates should have
- a strong background in Electronics/Electrical Engineering and should demonstrate experience of past work with inertial sensors
- (backgrounds in Computer Science, Robotics, Bioengineering, Mechanical Engineering are also suitable)
- must be familiar with common 3D representations, transformations, and reference systems
- experience with machine learning-based processing of (multi-modal) sensor streams
- programming experience, preferably in Python working with machine learning toolkits (e.g., TensorFlow, PyTorch)
- should have an expert understanding of underlying deep learning and machine learning concepts
- should be familiar with (generative) deep learning models (GANs, VAE, transformers, etc.)
- worked on signal processing as well as embedded systems before
(embedded work with IMUs, ideally experience with firmware as well as hardware prototyping)
Candidates should include a link to their portfolio, such as
- a website showing past projects (example1, example2) or
- link to a Github profile, blog, etc.
Requirements for application
- written and spoken fluency in English
- an excellent master’s degree (M.Sc., M.Eng. or equivalent) in Computer Science, Electrical Engineering, Bioengineering, or closely related
- experience working with large multi-modal datasets (e.g., activity recordings, physiology, multi-modal signals)
- experience implementing research prototypes (frontend, backend, or both)
- strong interpersonal and communication skills
Prior experience in conducting research and experiments using human-centered approaches in the fields of HCI, machine learning, computer vision, or biomedical engineering are a plus.