Subtle Presence: Design and Implementation of User Centric Content Delivery Using Biometric Data Capture and Intelligent Analysis
Symposium:
- ISEA2011: 17th International Symposium on Electronic Art
-
More presentations from ISEA2011:
Session Title:
- The Uses of Digital Database
Presentation Title:
- Subtle Presence: Design and Implementation of User Centric Content Delivery Using Biometric Data Capture and Intelligent Analysis
Presenter(s):
Venue(s):
Abstract:
This paper explores design and implementation of user centric content delivery using biometric data capture and intelligent analysis to determine the imagery and content presented.
Forms of data, captured in non-invasive manner, such as facial data, height, weight, body type, age, gender and aspects of mood can be used to alter content presented. Such systems must have an inherent intelligence that is ambient and ubiquitous – allowing for interpretation of a wide variety of stimuli that is easily captured. The systems intelligence must offer a range of options that can be autonomously responsive and give meaningful responses to visual and sensor cues.
There are many potential applications, information delivery for target advertising, social communication and emergency communications needed in public or social environments. Application can be used to create socially engaging artworks, integrated media delivery within architectural spaces and interactive media within an exhibition spaces. This allows for viewers engagement in aesthetic experiences that are subtly responsive to personal physical attributes and moods.
Technically, the design of this systems use inherent ambient, and ubiquitous intelligence through three model: Detection Model, Data Training Model, and Demo Showing Model. The Detection model algorithm detect a face, calibrates the image and extracts features using OpenCV Haar-like application (Viola & Jones) and LibSVM to classify and determine gender (SVM). AdaBoost learning algorithm is used to boost the classification performance. The Data Training Model uses classification method, LibSVM data file to train analysis of data and generate a final data model file. The Demo Showing Model manages windows for system and audience. The detection result data is shown in face detection window and scene view window. The content images or steaming video is shown in a second projection display.
This paper describes a system that demonstrates feasibility and successful application of responsive information delivery tools that prefigure the use of facial and biometric data to cue for advertising, social communication or delivery of culturally relevant user experiences. While initially designed for marketing content in public spaces, the content can vary depending on installation location, expected crowd and population demographics.
