Presence and Immersion. Ruth Aylett

Transcription

1 Presence and Immersion Ruth Aylett

2 Overview Concepts Presence Immersion Engagement social presence Measuring presence Experiments

3 Presence A subjective state The sensation of being physically present in the scene Rather than in actual location Aspects: Attention Physical or emotional reactions to events in the space. Memories of events in the space. 3

4 Immersion A physical state Being isolated from the outside world in a computer-generated space. Forgetting the outside world is there Provide computer-generated sensation to one or more of the human senses. Degree of control How far captures attention Visual Auditory Olfactory Haptic Taste 4

5 Rate of Immersion Immersion can take place slowly when watching a film the storyline needs some understanding before you can become engrossed A very fast rate of immersion can be achieved when a camera is on the front of a roller coaster The viewer experiences a plunging effect as the ride dips Activates the reflexive system Big changes in sensory input High degree of control

6 Rate of Immersion Therefore, a visual cue that improves immersion is a powerful way to gain a sense of presence in a virtual environment The quicker the immersion, the more a person feels engrossed Extended immersion time leads to adaption and a sense of complete presence

7 Factors Inhibiting Presence The sensation of presence can break down when the user is tired the weight of an HMD becomes uncomfortable or heavy for the user unnatural movement or lags occur during the experience

8 Engagement The state of caring more about the VR world than just it being a VR world The tasks become important Forgetting it s a simulation An affective state Enhanced by: Strong sense of immersion Well-designed tasks e.g. intuitive navigation Autonomous characters

9 Co-presence Feeling present with others in the virtual environment You and they share the same location Ability to make common references to the virtual world it s behind you ; over there

10 Social presence Not the same as co-presence Feeling that others in the environment are real people Their opinion counts You have a reputation to uphold You have a history with them Does not depend on physical presence Has been documented in text-only multi-user environments such as chat rooms

11 Open Questions Is there a definition of presence that is sufficiently operational and quantitative to be useful? What are the factors that create a sense of presence? Are there subjective and objective measures that can quantify presence? 11

12 Open Questions (cont.) Are there applications for which a sense of presence actually improves operator performance? Are there applications for which presence is a necessary ingredient? If so, how are these applications different from applications for which a more traditional display system is just as effective? 12

13 Achieving a strong sense of presence In general: Transparency of technology Sensory fidelity Good graphics especially depth cues Seeing parts of your own body High resolution and large field of view Familiarity of the VE or scene Binaural sound

14 Dynamic aspects Good interaction Amount: what can be changed Quality: whole body, unencumbered Responsiveness Absence of lags Maintaining convincing stereo vision and hearing as user moves

15 Achieving a strong sense of presence Seeing parts of your own body There is no doubt that this reinforces the feeling of presence Seeing computer generated parts of your own body tend to lower the feeling of presence Seeing actual parts of your own body tend to produce a strong feeling of presence

16 Achieving a strong sense of presence High resolution and large field of view Presence is reduced: If the edge of the image is visible If the field of view is restricted Ideally the user should just be able to move their eyes to see objects just outside the region of peripheral vision Engagement of peripheral vision has strong sensory consequences

17 Achieving a strong sense of presence Familiarity of the VE or scene If the VE is one that relates to a real world then the time taken to adapt to it appears to be shorter However, a user s experience may prevent them from becoming fully immersed because of artefacts in the VE

18 Measuring Presence Requires understanding of: the way a human interacts with the real world the adaptation that takes place when things change in the real or external environment

19 [Ellis, 1991] Our knowledge is constantly being updated by behavioural plasticity of visual-motor coordination and vestibular reflexes Thus, a large part of our sense of physical reality is a consequence of internal processing rather than being something that is developed only from the immediate sensory information we receive

20 [Sheridan, 1992] Presence is a subjective manifestation, much like mental workload and mental model - it is a mental manifestation, not so amenable to objective physiological definition and measurement Presence is a subjective sensation and as such any subjective measures are likely to be multidimensional

21 The aim.. Ideally we would like a set of repeatable objective measures for presence that indicate the degree of presence created by a particular system

22 How to measure Presence? Subjective measures Psychophysical measures Objective measures 22

23 Subjective measures To what extent did you experience a sense of being really there inside the virtual environment? A little A lot How realistic was your interaction with the virtual objects? A little A lot

24 Widely Used Subjective Presence Measures Steed, Usoh, Slater (SUS) Presence Questionnaire (M. Usoh, E. Catena, S. Arman, M. Slater, Using Presence Questionnaires in Reality. Presence, 2000, 9(5), ). Witmer and Singer (WS) Presence Questionnaire (B.G. Witmer, M.J. Singer, Measuring Presence in Virtual environments: A Presence Questionnaire, Presence, 1998, 7(3), ), Has been shown to be less effective in assessing the sense of Presence in virtual environments as opposed to a real world experiences, see: 24

25 Psychophysical measures Generally, psychophysical techniques are used to relate the physical magnitude of a stimulus with the observer s subjective rating of the stimulus magnitude. Example: R = f(s) where R is 1-7 feeling of being present and S is a screen resolution or lag time. 25

26 Objective measures Physiological measures Performance measures 26

27 Physiological measures Just as humans experience changes in physiological parameters in response to novel or unusual stimuli in the real world, given sufficiently realistic stimuli in a virtual environment, the human should experience similar physiological changes. Cardiovascular, Respiratory, Nervous, Sensory, Blood Chemistry 27

28 Performance measures Behavior Suspension of belief Ducking Socially conditioned reactions 28

29 The cliff experiment Subjects put into a virtual room Large hole or pit in the floor Asked to move their virtual representation around the room Very reluctant to walk over the pit

30 Zeltzer s AIP Cube One solution for characterising VE s was proposed by Zeltzer Assumes any VE has 3 components A set of models/objects or processes A means of modifying the states of these models A range of sensory modalities to allow the user to experience the VE

31 Zeltzer represents these components on a unit cube with vectors relating to Autonomy Interaction Presence (0,1,1) Ideal VE (1,1,1) (1,0,0) (0,1,1) Autonomy (0,0,0) (0,0,1) Presence Interaction CAD (0,1,0) (0,1,1)

32 Zeltzers Autonomy, Presence, and Interaction Cube (1,0,1) virtual reality (1,1,1) (1,0,0) (1,1,0) autonomy (0,0,1) (0,1,1) presence (0,0,0) interaction (0,1,0) We would not be able to distinguish between the virtual environment and the real world

33 Zeltzer s AIP Cube Autonomy A qualitative measure of a virtual object s ability to react to events and stimuli For no reaction, autonomy = 0 For fully comprehensive reaction, autonomy = 1 This scaling between 0 and 1 is purely qualitative

34 Zeltzer s AIP Cube Interaction The degree of access to the parameters or variables of an object For no real time control of variables, interaction = 0 For real time control of variables, interaction = 1 Whilst modern VE systems are capable of a high degree of interaction, the complexity of an application may inhibit or prevent interaction

35 Zeltzer s AIP Cube Presence A crude measure of the fidelity of the sensory input and output channels It is affected by the application of the VE

36 Zeltzer s AIP Cube The point (0,0,0) represents very early graphics systems programmed in non-real-time batch mode no interactivity Diagonally opposite, at (1,1,1) is ideal virtual reality maximum autonomy, interaction and presence so good that you wouldn t realise it wasn t real

37 Zeltzer s AIP Cube The point (0,1,0) can sometimes be achieved today user can control all the variables of some objects in real time The point (0,1,1) represents experiencing a high degree of interactivity and presence Some environments support regions close to this many VE s lack autonomy, though this is changing with the increasing use of physically based models and autonomous agents

38 Zeltzer s AIP Cube The point (1,0,1) represents a high degree of presence and autonomy a VE where the viewer is a passive observer but is fully immersed may be able to modify the viewpoint, but no objects should respond to a change in the viewpoint an IMAX film with stereo glasses & sound could possibly be considered here

39 Determinants of Presence Sheridan and Kalawsky proposed a total of 4 determinants of presence Extent of sensory information Ability of the viewer to change their viewpoint (for visual parallax or visual field) Ability to modify the spatial relationships of objects in a VE Closed loop performance due to a user induced motor movement (also includes the natural dynamic behaviour of moveable objects in the VE)

40 Determinants of Presence There is a temptation to concentrate on stimulating the senses rich 3D photo realistic display rich 3D soundscape tactile & force feedback but matching hardware interfaces to human sensory system is difficult, so we need to consider trade-offs

41 The sculpture experiment Constructing 3d rod-based sculptures

42 The aim Is it quicker/easier: From a workstation screen From a 3D VR representation Experiment 1 (1999) Physical example v Graphics display v HMD HMD and graphics display equal Low resolution of HMD Eyestrain and nausea One subject fainted

43 Experiment 2 Various setups in a CAVE Stereo v mono with shutter glasses Small-scale v immersive sculpture Tracking v joystick navigation

44 Results Life-size better than super-scale Super-scale produced longer times And more errors Head-tracking better than joystick use Joystick 20-40% longer

45 Phobias and Presence Specific phobias fear of heights, flying, bridges, post-traumatic stress requires presence Social phobias extreme shyness, fear of public speaking, stage fright, social anxiety, post-traumatic stress requires co-presence

46 Fear of Public Speaking Study Mel Slater, UCL FOPS an interpersonal phobia part of the general class social phobia Social phobia A persistent fear of one or more situations... in which the person is exposed to possible scrutiny by others and fears that he or she may do something or act in a way that will be humiliating or embarrassing. American Psychiatric Association manual DSM-IV

47 Co-Presence Requirement To elicit social phobic response VE must maintain illusion that there are other intelligent beings there capable of observing and reacting to the individual Our major question: To what extent will speakers react with appropriate affect to an entirely virtual audience?

48 FOPS Experiment (Summer 98 - D-P. Pertaub) 10 subjects rehearse in front of a virtual audience Talk repeated 3 times Two factor design Immersion HMD, desktop Audience response positive/negative audience (talks 1 and 2) negative through positive (ethical conclusion)

49

50 Audience response 5 subjects - first negative 5 subjects - first positive by 5 subjects immersed 5 desktop display Third talk Each subject concluded with a standing ovation

51 Avatar Design Avatars are moving, blinking, fidgeting, expressive Aim is portrayal of emotions expressionism rather than realism Two generations of avatar design first experiment stage fright scene + 2nd planned exp.

52 Response Variables Self-Rating subjects' evaluation of their performance. How would you rate your own performance in the talk you have just given? Assign to yourself a score out of 100, where 0 = completely dissatisfied with your performance, and 100 = completely satisfied.

53 Response Variables Modified PRCPS Personal Report Confidence as a Public Speaker I was in constant fear of forgetting my speech At the conclusion of the speech I felt I d had a pleasant experience I had no fear of facing the audience I felt disgusted with myself after giving my presentation I had no fear of facing the audience Count out of 16 such (negative) statements

54 Explanatory Variables Background age, gender, status. Ages 20s or 30s. There were 7 post-graduate students, 1 undergraduate, and 2 faculty members. None of the subjects were from Computer Science all were unknown to the experimenters before the study.

55 Co-Presence Explanatory Variables extent of being with a real audience 4 questions 1-7 scale In the last presentation to what extent did you have a sense that there was an audience there in front of you? To what extent did you have a sense of giving a talk to people? When you think back about your last experience, do you remember this as more like just talking to a computer or communicating to an audience? To what extent were you aware of the audience in front of you? Co-presence measured as no. of 6-7 responses/4

56 Explanatory Variables Perceived Audience Response subject's own impressions of the audience behaviour 3 questions, each on a 1 to 7 scale: How would you characterise the prevailing mood of the audience? How friendly were they? How would you characterise the interest of the audience in what you had to say? What sort of impression do you consider that you made on the audience?

57 Perceived Audience Response Independently of order there was a significant difference in 'perceived audience response'. 'good audience' (4.7 ± 1.2) 'bad' audience (2.3 ± 1.0).

58 Self-Rating by Avatar Interest rating rating

59 Self-Rating Regression Non-Immersed Immersed Negative Audience 21 5*C + 13*I 21 19*C + 13*I Positive Audience 73 5*C *C C = co-presence I = perceived audience interest R-Squared = 0.89 on 8 d.f.

60 Self Rating for Negative Audience Self Rating Co-Presence Non-Immersed Immersed For a fixed level of perceived audience interest immersion interacts with co-presence on self-rating.

61 Self-Rating for Positive Audience Self Rating Co-Presence Non-Immersed Immersed Perceived audience interest not-significant. Co-Presence amplifies response for immersed subjects

62 Self-Rating Conclusions Lowest self-rating:- negative audience, immersion, maximum copresence, and minimum perceived audience interest. Highest self-rating:- Negative audience: lowest co-presence and highest perceived interest Positive audience: highest co-presence

63 Modified PRCPS Logistic Regression Response = modified personal report of confidence as public speaker (high score means less confidence) Non-Immersed Immersed Negative Audience *I *I Positive Audience *I 0.8*C *I 0.8*C Co-Presence amplifies response for Positive Audience Chi-Squared = 12.2 on 12 df

64 Strong Observed Responses Clear (unmeasured) physical responses avoiding eye contact with experimenters skin colour and postural changes Spontaneous verbal statements I really bored them to tears. I m sorry about last time, I ll try to improve this time. Wake up! I see that my topic of the fetal heart doesn t interest you. That's a tough audience. I came out and my hands were sweating. Oh, how rude! Thank you, thank you very very much!

65 Co-Presence and Immersion No significant difference in reported copresence between immersed and nonimmersed subjects. A familiar result from our previous studies. Utility of Questionnaires???

66 Questionnaire Study Recent study (Martin Usoh++) Between-subjects, 2 groups 10 subjects per group Experimental group - immersed (HMD) Control group - in real world Both groups carry out search task Environments the same.

67 Conclusions When designing VE s we must look to human performance metrics and the subjective experience of presence to provide a measure of the effectiveness Presence and immersion play a very important role in VE s as they provide the only means of assessing the performance benefits of these systems

68 Credits Antony Steed, UCL Mel Slater, UCL William Winn, University of Washington Doron Friedmann, IDC, Herzilya