QML-CVC Research topics

The research we carry out represents a new area of work at the Computer Vision Center and thus its character is exploratory. The team explores the role of QML in topics such as:  

Generative models

We are interested in understanding the representation power of quantum systems to sample from distributions capturing relevant features to process stream audivisual data (e.g. musical concert, video-conferences). In this line, we are currently exploring potential computational advantages regarding resource consumption.

Image processing

We are interested in identifying the specific set of image processing problems which may be particularly suited to be solved with a quantum device. 

Graph networks


We are interested in exploring the potential resolution of bottlenecks related to dimensional size in graph processing, as well as generalization capabilities of quantum systems. The research question would be: Can the quantum circuits, thought as machine learning models, generalize from a graph structure?

Reinforcement learning


Quantum phenomena can be used to improve memory requirements to model the surrounding environment. Particularly, we are interested in quantum enhanced policy design in domain adaptation problems.

On the other hand we care about the capabilities of CV systems (e.g. quantum channel discrimination as a pattern recognition tool).  Which opportunities are offered by quantum correlations? 

ML Scientific discovery


While our universe is intrinsically quantum, all machine-learning algorithms describing governing dynamics work at the classical level. Here, we aim to expand the scope of such a field by processing data coming from a quantum sensor.