LifeGrabber
(March 2003)

Luigi Pagliarini<sup>1,2,3

1</sup>Maersk Mc-Kinney Moller Institute for Production Technology
University of Southern Denmark, Campusvej 55, 5230 Odense M., Denmark

² Academy of Fine Arts,
via di Ripetta 222, 00186 Rome, Italy

³ Laboratory of Cognitive Technology
Department of Psychology, Second University of Naples, Italy

e-mail: luigi@artificialia.com
http://www.artificialia.com/luigi

"LifeGrabber"(March 2003) is an Electronic Art piece consisting of a multimedia interactive installation that materialise the author's continuous research for new, computer and AI based, expression languages. "LifeGrabber", technically speaking, is composed by a PC, a Microphone, a WebCam located on a Robot Arm, the "ArmCam" (Pagliarini Jan. 2003, see Figure 1), and an AI software, "Cyberty" (Pagliarini Feb. 2003), that coordinates all of them in real time.

Figure 1
LifeGrabber - the "ArmCam"
(Copyright: Pagliarini 2003)

The mechanism in which it operates is relatively simple. Cyberty, is an AI based software that is able to analyse, real-time, input coming from both the audio and the video board. In terms of audio, Cyberty (that is a derivation of a former author' study, CyberInfinity) can use as input sound waves coming from a PC microphone or, alternatively, directly from a sound file. At the very same time, the software is able to capture, every now and then, images coming from a WebCam. In particular, in the "LifeGrabber" set-up the WebCam is located in the "hand" of a robot arm, which can be moved by the users (or spectators) or by the software itself. Once the full set-up is ready, the software starts collecting sounds through the microphone - or starts a musical files sequence - and initiate collecting images from the surrounding reality, synchronously. As an output, Cyberty produces images that can be either drawn on a computer screen or projected on a wall (see Figure 2 for an example). The images produced by Cyberty, obviously, depend upon what the computer set-up hears and sees. At the basis of the software there is a population of agents, conceived in an Alife fashion technique [1,2,3], that move on the screen. Each agent consists of a square with given dimensions and motion. Beside this, both the dimensions and the motion system are influenced, run time, by information coming from the audio board. In the very same way, each agent, besides of a proper colour, has an image filter that uses to interpolate information coming from the WebCam. Therefore, each cycle of the algorithm, once agents have received info from the audio-visual system, they modify themselves according to it and, after that, they are drawn on the screen. In this way, for each screenshot, the resulting picture is at the same time a track of present, past, and future. Indeed, some of the audio information, like music waves frequency and amplitude, might elaborated in advance (i.e.: only when the software deals with audio input coming from a sund file).

Figure 2
LifeGrabber - "My Eye" - Screenshot
(Copyright: Pagliarini 2003)

Moreover, while moving, agents when they meet they influences each others by opening a discussion about colors, dimensions and etc [4]. This "social" mechanism ensure that the resulting final image expressed on the screen does never reach saturation and is always balanced in colors and other graphical characteristics. "LifeGrabber" set-up can vary a lot and well adapts to various architectural circumstances. The author's preferred is the one titled "Meditation" (see Figure 3 for an example) where the WebCam is directed to its own projection, therefore being engaged in an endless loop between what it sees and what it projects.

Figure 3
LifeGrabber - "Meditation" - Screenshot
(Copyright: Pagliarini 2003)

This "self observing machine" concept is to be considered the most interesting and futuristic one because, such sort of short-circuit, somehow concretises the long time followed human dream for the "machine self-consciousness". In short, this work is in the very same direction of Nam June Paik's "Buddha TV" (1989) , but moves a little further.
Mostly, "LifeGrabber", as an art piece, frees itself from the author to give the clue of the aesthetic mean to the environment, here included observers (see [5] for further details).
The author is now working on a new "LifeGrabber" version where the multimedia installation produces, environment influenced, agent-based, sounds.

References
[1] - Langton, C.G. Artificial Life. In L. Nadel e D. Stein (ed.) Lectures in Complex System, SFI Studies in the Sciences of Complexity, Lect. Vol. IV, Reading MA, 1992.
[2] - Sims, K. Artificial Evolution for Computer Graphics. Computer Graphics 25, 4, 319-328. 1991
[3] - Lund, H. H., Pagliarini, L., and Miglino, O. Artistic Design with Genetic Algorithms and Neural Networks. In J. T. Alander (Ed.) Proceedings of 1NWGA, Vaasa University, Vaasa. (1995-a).
[4] - Spina, A., Pagliarini,L., Globalization: tra arte scienza e società. In Rivista di Psicologia dell'Arte Anno XII, n.12, 87-92. Roma, Dec. 2001
[5] - Pagliarini L., Locardi C., Vucic, V. Toward Alive Art. In Proceedings of Virtual Worlds 2000. Second International Conference, J.C. Heudin (Ed.) Springer-Verlag Press (2000).

Acknowledgements
The idea and the development of the "LifeGrabber" was, directly or undirectly, strongly influenced by the italian artists: mou, lips!, Jaromil and bianco-valente.

Video

Luigi Pagliarini is:
 - Professor of Machine Psychology at the Academy of Fine Art of Rome.
 - Professor of Techniques of Multimedia Programming at the Second University of Naples, Psychology Faculty.
 - Lecturer of Robots and Dynamic Systems Interfaces at the Maersk Institute, University of Southern Denmark.
 - Board Commitee of Rivista di Psicologia dell'Arte.
 - Executive Member of the Co-Operative Project Inserm (France) - CNR (Italy) for Biomedical Research.
 - Executive Member of the RoboCup Junior International Committee.
 - Member of the International Artistic Commission Vivaria.
 - Member and Collaborator of G.R.A.L. (Research Group on Artificial Life, C.N.R. - Italia).
 - Member of EvoNet (European Network on Evolutionary Computation).