GRAF brings you the programme of contemporary art in Catalonia. The agenda that combines all agendas.

Computer Vision

16:0018:00 h
Venue: Hamaca
Emilia Coranty, 16 - Can Ricart
08018 Barcelona
Price: Free event

Until recently Computer Scientists were considering the images as the “dark matter of the Internet”. Programmes were able to analyse colors, light, shapes, but were struggling to discover the semantic content of photographs and movies. For a computer, an image consists of a series of numbers, and to translate these numbers in a coherent description is a hard problem. Computer Vision is the name of the effort to translate the visual competencies of humans to machines. These last five years, this discipline has produced impressive results. Machines are increasingly making sense of visual content. In many contexts, they perform tasks informed by a fine grained analysis of images. They help doctors to diagnose various illnesses, they scrutinize the never-ending footage of surveillance cameras, they detect new planets from telescope imagery. When they assist car drivers, they recognize pedestrians and avoid obstacles. To achieve this, the current techniques of Computer Vision learn from humans how to look at images. Which means that lots of anonymous workers describe images for machines. And from these descriptions, the machines gradually produce a model of the world. And not another. This work of description of the visual world made for the machines is called annotation. The annotators spend their days pointing things in images and naming them. During these two days of workshops, we will occupy the position of these annotators to look freshly at images. We will explore what we can learn from images when we describe them for machines, as well as the political, economic and aesthetic implications of doing so.
To do this, we will make two experiments, one each day:

First one is inspired by “What do we perceive in a glance of a real-world scene?” Ese experimento fue imaginado en 2017 por Fei Fei Li en el laboratorio del California Institute of Technology para establecer una relacion entre retina, temporalidad, imagen y palabras que sirve de base para la visión por ordenador. During this session we will look at images during less than half a second and explore what we can see and say about images when we look at them at machine speed.

Second is inspired by “La habitacion china”, una situación relacional imaginada por el philosofo John Searle en 1980/4 que interrogó performáticamente la idea de inteligencia artificial, poniendo en tensión el binarismo máquina-humano. During this session, we will learn to describe images without seeing them, just like a machine.

During these two experiments, we will look at images, and these images will come from the video database of Apología/Antología. What we will see and what we will say about these images will serve as a starting point to see them differently and suggest different (perhaps inappropriate/d?) modes of description for them.