April Dances in Coimbra

April Dances in Coimbra

Cátia Roça

Filipe Amaral

Department of Informatics Engineering, University of Coimbra, PT

Autonomous poster design system

Keywords: Poster; Interactive; Computer Vision.

The vision of the future will certainly refer to the vision of machines, as mediators or autonomous agents. These machines, in addition to many other purposes, will be responsible for communicating information to humans more effectively. This project makes use of computer vision to create an autonomous machine (software) capable of personalizing the design of a poster to the person who is in front of it, as well as highlighting the message it aims to communicate.

The system was applied in the creation of a promotional poster (speculative) for the dance event "Abril Dança em Coimbra", in Coimbra, Portugal. Whenever a person stands in front of the poster, music begins to play and the person's movements are mirrored by coloured lines that make up the poster. Thus, the person is motivated to get into the spirit of the event and give life to the poster by dancing.

On the hardware side, this interactive digital poster is implemented using a screen, a computer and a camera. The software is developed in p5.js and ml5.js and make use of the neural network "PoseNet" to detect the body structure of the person ahead. In its default stage, the poster remains static. Whenever a person is detected, a randomly picked tune (free rights) starts to play and some coloured lines that make up the design are animated mirroring the person's moments. The design is re-colourized depending on the music's style, helping to bring the viewer into the mood of the event.

April Dances in Coimbra

The Future Vision Online Exhibition was available from 11 February to 11 March 2021.

What you find here is a fleeting glimpse of the 21 works – or in better words, 21 future visions – coming from artists, designers, researchers, professionals, and students who explored complex questions from a critical and creative perspective.

We extend our warmest thanks to everyone who participated in the PCD21.

View all works
FBAUP i2ADS