Embodied Self-Motion Illusions in VR

profile

How can we pro­vide humans with a believ­able sen­sa­tion of being in and moving through com­put­er-gen­er­at­ed envi­ron­ments (like VR, com­put­er games, or movies) with­out the need for costly and cum­ber­some motion plat­forms or large free-space walk­ing areas? That is, how can we “cheat intel­li­gent­ly” by pro­vid­ing a com­pelling, embod­ied self-motion illu­sion (“vec­tion”) with­out the need for full phys­i­cal motion?

In a series of exper­i­ments, we inves­ti­gat­ed the con­tri­bu­tion and inter­ac­tion of dif­fer­ent sen­so­ry modal­i­ties as well as higher-level, cog­ni­tive influences.

Below is a selec­tion of such work [more will come as the web­site evolves]

Spatialized Sound Enhances Biomechanically-Induced Self-Motion Illusion (Vection)

The use of vec­tion, the illu­sion of self-move­ment, has recent­ly been explored as a novel way to immerse observers in medi­at­ed envi­ron­ments through illu­so­ry yet com­pelling self-motion with­out phys­i­cal­ly moving. This pro­vides advan­tages over exist­ing sys­tems that employ costly, cum­ber­some, and poten­tial­ly haz­ardous motion plat­forms, which are often sur­pris­ing­ly inad­e­quate to pro­vide life-like motion expe­ri­ences. This study inves­ti­gates whether spa­tial­ized sound rotat­ing around the sta­tion­ary, blind­fold­ed lis­ten­er can facil­i­tate bio­me­chan­i­cal vec­tion, the illu­sion of self-rota­tion induced by step­ping along a rotat­ing floor plate. For the first time, inte­grat­ing simple audi­to­ry and bio­me­chan­i­cal cues for turn­ing in place evoked con­vinc­ing cir­cu­lar vec­tion. In an audi­to­ry base­line con­di­tion, par­tic­i­pants expe­ri­enced only spa­tial­ized audi­to­ry cues. In a purely bio­me­chan­i­cal con­di­tion, seated par­tic­i­pants stepped along side­ways on a rotat­ing plate while lis­ten­ing to mono mask­ing sounds. Scores of the bi-modal con­di­tion (bin­au­r­al bio­me­chan­i­cal cues) exceed­ed the sum of both single cue con­di­tions, which may imply super-addi­tive or syn­er­gis­tic effects.

Stepping along a rotat­ing cir­cu­lar tread­mill while being blind­fold­ed can (after a few sec­onds) create a com­pelling & embod­ied illu­sion of rotat­ing in the oppo­site direc­tion. Studying such “bio­mech­ni­cal­ly-induced cir­cu­lar vec­tion” can help us to  better under­stand human multi-modal infor­ma­tion pro­cess­ing and even­tu­al­ly help us to improve motion sim­u­la­tion in Virtual Reality.

 

First demo of our new cir­cu­lar tread­mill in the iSpace lab. Since then the setup has had some major updates.

 

Why not build an omni-direc­tion­al tread­mill so people can walk in all direc­tion in the Virtual Environment? While recent progress has been amaz­ing, such omni-direc­tion­al tread­mills are still huge, loud, expen­sive, and need many safety mea­sures (see video below for an exam­ple from the Cyberwalk project). In addi­tion, it seems that walk­ing-in-place can only reli­ably induce cir­cu­lar vec­tion, but not translational/linear vec­tion. Hence, we decid­ed to focus on using out under­stand­ing if human multi-modal cue inte­gra­tion to “cheat intel­li­gent­ly” and aiming for a com­pelling embod­ied illu­sion of self-motion (i.e., “per­cep­tu­al realism/effectiveness”) instead of aiming for biomechanical/stimulus realism.

Media Gallery

[mediatags]biomechanicalVection[/mediatags]

Related Publications

37904 vec­tion 1 apa 50 “year” 688 https://iat-ispace2.dcr.sfu.ca/wp-content/plugins/zotpress/