Abstract: Virtual, mixed, and augmented reality, collectively referred to here as extended reality (XR), provide users with powerful immersive visual interfaces that allow for intuitive view selection, for depth perception, and for the seamless blending of real and virtual environments. However, traditional immersive visual interfaces render all of the virtual and real environments strictly from the user’s viewpoint, which limits their efficiency and effectiveness. For example, a user’s view of a complex 3D virtual environment is limited by occlusions, and the user has to navigate the virtual environment to find a viewpoint that provides the desired line of sight. When an occluded region turns out to be of no interest, the navigation is wasted. As another example, in a collaborative XR application the multiple individual user viewpoints complicate cross-referencing the shared workspace. As a third example, requiring a strict alignment of a physical object with each virtual object with which the user intends to interact severely limits the opportunities for XR haptic feedback. In this talk, I will present a versatile approach for increasing the efficiency and effectiveness of XR visual interfaces, based on viewpoint generalization. The benefits of the approach will be demonstrated in a variety of XR contexts, including environment exploration, shared workspace collaboration, haptic feedback, simulated transparent displays, environment complexity management, and the fine grain interleaving of immersive and non-immersive visualization.