Student projects

Below are some project ideas for MSc students

1 – Autonomous drone navigation to and landing on user-specified visual markers

The student will use the available Phantom DJI 3 standard drone equipped with a video camera. The aim of the project is to develop a smart phone app whereby the user would be allowed to select among a set of visual markers, which will be systematically or randomly distributed in a field. In turn, the drone should be able to navigate to and land, as precisely as possible, on the selected marker.

2 – Projects on Internet of Things and Block Chain

You are welcome to come with your ideas for IoT- and Block chain-related projects

3 – Security biometric systems

Modern vision-based biometric systems, such non-invasive contactless fingerprint and palmprint recognition, are becoming important technologies for the industry. Ideas in these fields are welcome.

4 – Projects in collaboration with Crowdynews

Selected candidates in any of the projects below will have the opportunity to work with the AI team of Crowdynews in Groningen

a. Logo detection in images and/or videos – The aim of this project is to localize and recognize all logos of interest in given visual data. For evaluation, we will use a combination of benchmark data sets and data obtained from clients of Crowdynews

b. Automatic classification of videos as safe and not safe – The aim of this project is flag a given video clip of having safe or unsafe content. Examples of unsafe content may include scenes or objects related to terror, for instance.

c. Gender recognition from behavioural data on websites – The aim of this project is to infer the gender of the user accessing a given website. The hypothesis is that properties such as, duration, frequency and time (among others) may have sufficient discriminative power. For evaluation we would need to collect as much annotated data as possible. This means, we would need to invite many volunteers to interact with a given website and declare their gender.