PDF Merged Merged
PDF Merged Merged
PDF Merged Merged
190
120
It creates light into electrical signals. The image on the retina is upside down due to lighting
bending.
How many types of receptor cells do we have and where are they located?
Two: rods (which are used for peripheral vision B/W) cones (color), They are located in the
retina
What range of the electro magnetic spectrum is visible to the human eye?
380-700 nm
Water 99%
Red-Green-Blue
Name the three color channels based on three types of pigment code?
· Achromatic
· Red-Green Channels
· Blue-Yellow channels
Men’s because they have only 1 X-Chromosome while women have 2 and in order for them
to have color blindness both X-chromosomes must have a faulty gene
Name the three types of color blindness and analyze their disfunctions?
Key notions :
· what is the human binocular vision? (the ability to use information from both
sides at once)
· The eyes adapt to color intensity yielding color after effects.!!! (how do the eyes
adapt to colour intensity?)
Motion Vision
What are the differences between Parvo cells and magno cells?
Parvo cells: color sensitive, slow, high resolution. They can figure out shape and
color/luminance
Magno cells: color blind, fast, low resolution. They can figure out Motion/depth and shape
what can we say about local motion (downside) and object motion?
Local motion is ambiguous and must be further integrated into object motion.
Key notes:
· Local motion is ambiguous and must be further integrated into object motion
Stereo Vision is the ability to recognize the depth in which objects appear by comparing the
different of views of the same scene (left- right eye)
Stereo blindness is the inability to perceive stereoscopic depth by processing the disparities
between the images from both eyes.
Medical disorders
What are the similarities and differences between stereo and motion vision?
Stereo vision is similar to motion vision since they both need two different views of the same
scene. However stereo visions main difference is that depth requires the scene to be static
Key notes:
· How can we reconstruct depth from two different views? (we can do this as long as
we know the relative viewing angles of both eyes)
Paper 1 : Multiple sensors use in order to predict a cue. MLE model is used in order to
reduce the variance among the guessed outcomes. Reduce of variance was successful after
putting multiple sensors to test to predict 1 que (predict the size of an object using both
visual and haptic ques).
Paper 4:
· emotional responses to an environment are context dependent and not
dominated by a single sensory modality
· Multiple sensor modalitys that are well combined can evoke positive emotions
· Multiple sensor modalitys that are not well combined evoke negative effects
All in all the paper did not manage to make a firm conclusion on how different environmental
characteristics should be combined in order to enhance positive emotion.
Hearing
Frequency coding is the local resonance of the basilar membrane where specific sound
waves at specific frequency’s trigger specific locations on the basilar membrane.
The pitch of sound is determined by the frequency of vibration of the sound waves that
produce them. A high frequency (e.g., 880 Hz) is seen as a high pitch, while a low frequency
(e.g., 55 Hz) is regarded as a low pitch.
What is the difference and which are the similarities between harmonics 1 to 10 and
pure tones?
· Difference: timbre
What is the difference and what is the similarity between harmonics 4 to 10 and pure
tones?
In decibel
What is a decibel?
The decibel (dB) is a logarithmic unit that indicates the ration between:
What is stereo hearing (what are the parameters that influence horizontal and vertical
direction of sound)?
Which compress techniques are used by MP3 in order to reduce the data of a song?
· Joint stereo: often same information in left/right channels thus with joint stereo we
reduce the final size by using less bits for the side channel
· Huffman encoding: creates variable length codes on a whole number of bits. Most
frequently occurring information have shortest code. The decoding step is very fast thus
it allows to save 20% space on avarege
· Psycho-acoustic masking: this method filters out low amplitude sounds that are played
alongside with high amplitude sounds. When this occurs in a song the low amplitude
sounds can not be heard by the human ear thus this methods excludes them.
Key notes:
Touch
· Proprioception (the ability to be able to tell the relative position of body parts. For
example being able to touch the nose with your index finger while having eyes closed)
· Cutaneous sense: any sense that is dependent on receptors in the skin. For example
pain, temperature, vibration etc.
Yes
Key notes:
· Touch is important for motor coordination and pain warnings but can also be used to
communicate information
· Tactile interfaces can be used for spatial awareness, threat wanings, way finding and
crew communication
Multimodal Perception:
· Complementary (color and weight of an apple) where the sensors do not directly depend
on each other.
· Redundant (size and weight of an apple) where the different sensors present the same
information.
It weighs the individual sensory estimates such that the total variance is minimal.
YES! Human brain takes into consideration reliability and uncertainty of each sensory
modality. However, there are situations where this is not possible because there is a conflict
between informations provided by each sensor. In these cases the brain may prioritize one
source of sensory over the other.
· Sound alters visual percept (There was some example where two sound beaps made
people think that a light flashed twice)
· Sound influences motion direction (moving sound captures dynamic random dot pattern
(for more see documentation))
Key notes:
• Optimal sensory integration weighs the individual sensory estimates such that the total
variance is minimal.
• The integration of incongruent sensory information can lead to illusions (e.g. rubber
hand, flash-tap illusion, odor).
• Visual, vestibular and cognitive cues integrate into the percept of the subjective vertical.
What is synesthesia?
VR Technology:
Why is VR useful?
· Performance monitoring
· FUN
· Sensory feedback
· User control
Give the similarities and differences between natural vision, flat world vision,
Stereoscopic displays (VR).
Natural Vision 2D 3D 4D
Blur as depth cues
focus on either
object ü - - ü
Stereoscopic
disparities as depth
cues ü - ü ü
Vergence angle
and
accommodation ü - ü ü
convary and give
depth cues Fixed
accomodation
What are the two different displays of augmented reality? What are the advantages
and disadvantages of them?
· See-Through:
Advantages:
Disadvantages:
Advantages:
1. Color change (DanKam app smartphone)
Key notes:
· The interaction loop consists of : (head and hand) tracking, virtual world simulation,
sensory feedback and user control.
· Visual stereoscopic feedback requires the separation of left/right eye images through
color differences, polarization or multiplexing. (what is required in order for visual
stereoscopic feedback to work?)
· Veridical stereoscopic perception requires matching the simulated viewing position and
direction with the actual viewing position and direction (through head tracking, CAVE, or
HMD). (what is required for veridical stereoscopic perception?)
· Proprioceptive feedback (for vestibular organ and muscles) requires motion platforms
and force feedback devices. (what is required to achieve proprioceptive feedback?)
VR Applications:
· Training (military)
What are the challenges and solution of the planetary rover example (Telepresence)?
What do we need to take into account while developing a user interface for vr?
Navigation
· Navigation based on route representation: it involves sequences of local views that are
associated with certain actions (turn right)
· It is formed based on self localization: internal and external self motion cues can be used
to maintain a sense of position and orientation
· Based on spatial attributes: visual information about spatial attributes of the environment
and the objects therein.
Genders have their own way of navigating what is the main difference in navigation
strategies between males and females?
Men use geometric cues for navigation while females apply a landmark strategy (memory
holding using landmarks).
Are there multiple sensors used for navigation give an example where multisensory
navigation is better?
Yes (example: experiment with virtual mazes where Audio Visual landmark had better results
in terms of spatial memory and navigation performance in compare to just Visual or just
Audio)
Key notes:
When a robot is made more humanlike in its appearance and motion the emotional response
from a human being to the robot will become increasingly positive and empathetic. However
when a certain point is reached going beyond that will create strong repulsion.
Give the model for affective appraisal?
What is embodiment?
· Physical Presence (being in another physical world then where your body is
located)
· Body ownership
· Self-location (sensation that locations of you and your body coincide in space)
· Questionaires
· Self-location or proprioceptive drift (perceived position of own (hidden) hand versus drifts
towards rubber hand)
· Anthropomorphism
Key notes:
• The Uncanny Valley : the bandwidth of near-realistic humans can yield strong repulsion.
· Invasive BCI(Electrocorticography)
· Writing the visual cortex (a way to decode picture so that blind people can see)
Key notes:
• BCIs are (non)invasive systems that measure CNS activity and convert it into artificial
output that replace, restore, enhance, supplement, or improve CNS output. (what are
BCIs?)
• Today, most promising for use in (mobile) human computer interaction are non-invasive
EEG based BCIs, which have high temporal resolution (good for control) but low spatial
resolutions (still hard to extract meaningful signals).
• Examples of reading the brain (non-invasive): attentional state, emotions, what someone
sees, (motoric) intentions.
Cyborgs
See prosthetic memory!!
Key notes:
Google glass / Hololens: Augmented reality with internet connectivity. Far future: visual
prosthesis?
Sharing through networks (BCI, brain-brain): team members tap each other’s sensors and
experiences.
Lecture7
· Perception (recognition)
· Communication (verbal)
· Bodyparts (mostly hand tracking) (also called natural interaction) (gorilla arm effect
downside – sometimes its hard to percept depth of object- does not feel real) (pros: using
gestures allows you to reach object far away)
· Natural devices (physical form that translates to digital information. Example paddle for
operations on a catalog book)
Give pros and cons of special devices and natural hand tracking methods in AR?
Special devices: + tactile sensation + accurate, -special hardware needed (unutural),
-difference between interacting with real and virtual objects
Natural hand tracking: + feels real, + user can interact with virtual object everywhere in
space, - usually no tactile feed back, -memorization, feedback,accuracy (for gesture
interaction)
· Distracting,overwhelming
· Performance
What do we need to take under consideration when creating interaction design for AR?
There is no one ideal interaction. However, usually the best approach is for the design to feel
natural.
· Restricted FOV
· Luminance mismatch
· Contrast mismatch
· Position
· Rotation
· Scaling
Give some selection issues in AR interaction?
· Non-isomorphic (Magic virtual tools that extend the working volume of the
arms(application dependant))
· Object size
· Ray casting
· Cone casting
· Egocentric metaphor (Ray casting, aperture etc. This are interactions that happen within
the virtual world)
· Exocentric metaphor (world in miniature. Here you basically get a miniature world of the
virtual world (within the virtual enviroment) and by interacting with it you make changes
to the virtual world (gives more control)).
· Precision placement
· Amount of rotation
Key notes:
EXAM BASED
Different input device categories (know them including TUIs and characteristics, be
able to make relations between technology aspects & usage (sensors for tracking &
how they are used for interaction))
Interaction design & tasks (comparison of characteristics, pros and cons, relation
between technology and possible usage)
LECTURE 6
● Field of view
Extra hardware and computation to get user’s perspective on screen cause of camera lens
and distortion.
● Eye focus
The human eye can focus only at a certain depth of our vision.
Oculomotor cues:
● Accommodation
● Convergence
See-through HMDs
AR head mounted displays. Basically AR glasses. He goes over the problem of depth
focus, since we project on just a single display and how a work around that they are looking
into now is a specific camera that when you take a picture, it takes multiple 'versions' for all
depths (by recording all the light that comes towards it from all angles?). You can then use
that alongside eye trackers to automatically adjust what should be in focus in real time, with
better hardware.
Lecture5
Give AR characteristics?
Hand held
Spatial
Head worn
Multimodality?
Application/usage?
Multi-users
Give an example where for two different systems that are at the same spot on the
continuum the user experience would differ a lot?
Head mounted display, handheld display with same visuals and audio and everything exactly
the same. However, phone vs head mounted display are completely different in terms of
user experience.
Exam based: know what type of AR we can create with each incarnation?
What is the AR interaction loop?
VR/AR comparison :
Key notes:
AR: know what to use when and what you can or cant do given certain circumstances and
goals
Which characteristics do sensors have that are used for tracking in AR?
Measures rotation rate of rotation better than accelerometers but drift over time.
(((((Marker tracking usually uses black and white features cause of contrast () while in
natural tracking you need texture and when some amount of the same texture appears over
the screen the object is recognized as a flat surface ))))))
They don’t adapt well in size/distance and angle changes and they have high latency.
Inertial (gyroscope,accelometer)
Optical (marker based, natural feature based)
Structured light cameras
others