MOVING OBJECTS IN SPACE :
EXPLOITING PROPRIOCEPTION IN VIRTUAL ENVIRONMENT INTERACTION



By

Mark R. Mine
&
Fredrick P. Brooks Jr.

University of North Carolina at Chapel Hill

Carlo H. Sequin

University of California at Berkeley






Overview

What is Proprioception ?

Problems in Interaction in Virtual Environment



Using Proprioception.



Interaction Techniques

1. Direct Manipulations



2. Physical Mnemonics


(a) Pull Down Menus

(b) Hand held Widgets

Hand Held Widgets are preferred over the Object bound widgets due to :

(c) FOV - Relative mode switching

(3) Gestural Actions

(1) Head Butt Zoom -



The advantages this method provides :




(2) Look-at menus



Besides being intuitive ,gives the user the abilty to use the input device for other things , while making a selection in this menu.


(3) Two-handed flying





(4) Over the shoulder deletion



Its utility could be explained in terms of ease of remembering the metaphor, and not using any menu space or buttons .




User Studies

(1) Virtual Object Docking

Goal :

Task :

Target shapes transparent and red in color were to be aligned with opaque blue colored cubes. These blue cubes were either located at User's hand position or at a distance, and then the time taken by the user to align the two cubes was noted.

Each user did 216 trials , ie 36 docking trials /test X 3 tests/block X 2 Blocks

Results

Experiment Results

Questionaire Results

Note : The points are given on the scale of -3 of 3


(2) Virtual Widget Interaction

Goal :

Task :

The Users were expected to carry out a 3 step procedure , once for the hand held widget and then for the widget which was floating in the air , at a fixed distance.

In the case of the hand held widget , the widget moves along with the users hands. In the other case they remain fixed irrespective of the users movements. For both the cases , the Users were supposed to first align a 3D Cursor to one of the 6 spheres on the widget , then align it to an arbitrary object in the space , and finally bring it back to align it with the widget , without any visual feedback.

The positional accuracy of the user , in aligning the cursor was noted , and evaluated.

Results:

Experimental Results :

The following table shows the mean positional accuracies in the 2 cases , and the results clearly show , that the hand held widget gave more accuracy in alignment.

Questionaire Results:

These results also indicate the relative preference for Hand Held widget System rather than a widget system away from the body.