[ Content | View menu ]

description of video content for the flashlight matrix

Written on September 25, 2008

Recent developments in computer-aided surveillance technology suggest artificial intelligence algorithms that are based less on the comparison of facial features than the analysis of  movement patterns. For example, the department of electrical and computer engineering of the University of Maryland’s A. James Clark School of Engineering has developed a surveillance system that analyses human gait and corresponding activities, such as humans carrying objects like backpacks, handbags, or briefcases for  “suspicious activities”.

This is the conceptual framework for a dance performance in which the performers develop an abstract choreography based on the afore mentioned technological concept of suspicious movements. The choreography is designed in a non-linear fashion, individual parts will be recombined by computer software (click on the image below to see a higher resolution diagram):

We will tape the performers with surveillance cameras to create a high contrast between the background (black) and performer (white). This footage is the main source for the 10×10 flashlight matrix. At the same time, using pattern matching algorithms, certain movements in the recorded dance performance trigger related videos in a database of prerecorded film excerpts. After the film excerpt is played back, the video switches back to the dance performance.

We are deliberately taking into consideration the subtle glitches and errors of our pattern matching algorithm that will create new and unexpected combinations of the combined materials.

Filed in: Development.

No Comments

Write comment - TrackBack - RSS Comments

You have to be logged in to post a comment.