Visual Manipulations for Improved Generalized Presentations

In many cooperative work sessions a presenter shares a view of her workstation screen with a group of passive viewers, projecting it on a large screen or using bitmapbased sharing protocols. While these schemes offer advantages of simplicity and sharing transparency over custom collaboration-aware tools, all group members will have the exact same view. Data deemed private by the presenter may be exposed and viewers have to watch the presenter’s detailed interactions, making it hard to infer her underlying intentions. We present a general system that allows customization of existing single user application views. Simple policy-guided manipulations are applied to the shared windows and image buffers, providing proactive privacy protection with relaxed verbosity and cues to meet presenter and viewers' needs.