Metropolis Crowd ControlWhen does a crowd grab your attention? Can you replicate the unique ways people move? What does a crowd sound like? Can you tell the mood of a crowd?

Metropolis is a novel research project, which aims to answer these questions by combining expertise in computer graphics, engineering and cognitive neuroscience to create highly realistic virtual crowds.

At METROPOLIS: CROWD CONTROL you can check out the demos of the research carried out so far and take part in real scientific studies conducted by Metropolis researchers. Your participation will help us answer more questions about virtual crowds and contribute to the development of the computer games industry, urban planning, pedestrian and traffic modelling, evacuation simulation and assistive technologies.

Find our more about the experiments you can take part in below.

Metropolis is supported by Science Foundation Ireland.

Prof Carol O' Sullivan,  Prof Fiona Newell, Prof Henry Rice

Trinity College Dublin

Find out more at GV2


When we look at an angry mob or a group of football fans, we can easily sense the overall emotion of the crowd. In this study we are looking at how accurately we can rate crowds in terms of their emotions.


How quickly can you spot a particular person amongst a crowd of people? Can your other senses help you look?  In this experiment, you will try to find your friend in a crowd of people accompanied by various sounds.


Can you pick someone out of a crowd just from their voice? Our brain uses binaural (both ears) cues to figure out positional information about what we hear. This experiment will help determine how good we are at localising the position of a sound.


Could you estimate how many people were at St Patrick's Festival or the Dublin City Marathon from just looking? Help us find out the point at which a group becomes a crowd by guessing the number of people in our virtual crowds.


If we see others looking in a particular direction we tend to look that way too. We are investigating how many people in a crowd are necessary to look in a particular direction before we follow their gaze.


Traffic is something we're all too familiar with and deal with on a daily basis. In this experiment participants will see how quickly and accurately they can locate a moving object in a busy traffic scene.


We can recognise our friends or ourselves from dot patterns of walking gaits alone. But, can we tell whether the motions of virtual humans were captured from real people, or if they were the synthetic creations of a graphics wizard?


Have you ever seen a real crowd full of identical clones? Usually, in the real world everyone around us looks unique and we are very good at recognising people from their individual appearance. In this experiment participants will help researchers find out what changes make people in a virtual crowd look as different as possible.


Could you spot a wolf in sheep's clothing? To make a virtual human crowd believable you need lots of variety. But what about herds of animated animals - do you need the same amount of variety? Watch the motions of our virtual farm animals and tell us whether you think the animals are natural or not.