Collaboration: TU/e, Philips
Duration: September 2017 until September 2021
Contact: Milan Petkovic
The topic of this PhD focuses on the study, design and development of skin specific descriptors that allows for creation of tailored analytics software to be used for automatic detection of skin features. This will lead to direct products application primarily for personal care and potentially for skin health.
Within the four years of PhD, skin images from various device modalities will be acquired in the context of multiple studies in beauty, personal health and health care. The successful candidate will primarily focus on understanding, using and tailoring computer vision tools in order to parametrize skin images.
The combination of parametrized and annotated data will be used as a main input for machine learning algorithms to create skin-specific features detectors. The candidate is expected to use current machine learning techniques ranging from regular classifiers to deep convolutional neural networks, improve them for skin-specific applications as well as develop new modalities when applicable.
On a technological level, the candidate will develop expertize in computer vision and related machine learning techniques.