universal design principles - NCSW • equitable use chapter 10 • flexibility in use • simple and intuitive to use universal design • perceptible information • tolerance for error • low physical effort • size and space for approach and use Multi-Sensory Systems Usable Senses The 5 senses (sight, sound, touch, taste and sm ell) are • More than one sensory channel in interaction used by us every day – e.g. sounds, text, hypertext, animation, video, gestures, – each is important on its own vision – together, they provide a fuller interaction with the natural world • Used in a range of applications: – particularly good for users with special needs, and virtual Com puters rarely offer such a rich interaction reality Can we use all the available senses? Will cover • – ideally, yes – general terminology – practically – no – speech – non-speech sounds We can use • sight • sound • touch (som etim es) – handwriting • considering applications as well as principles We cannot (yet) use • taste • sm ell Multi-modal vs. Multi-media Speech • Multi-modal systems – use m ore than one sense (or m ode ) of interaction Human beings have a great and natural e.g. visual and aural senses: a text processor may speak the mastery of speech words as well as echoing them to the screen • Multi-m edia system s – m akes it difficult to appreciate the – use a num ber of different m edia to com m unicate complexities inform ation but e.g. a computer-based teaching system: may use video, – it’s an easy m edium for com m unication animation, text and still images: different media all using the visual mode of interaction; may also use sounds, both speech and non-speech: two more media, now using a different mode 1
Structure of Speech Speech (cont’d) phonemes Other term inology: – 40 of them • prosody – basic atom ic units – alteration in tone and quality – sound slightly different depending on the context – variations in em phasis, stress, pauses and pitch they are in, these larger units are … – im part m ore m eaning to sentences. allophones • co-articulation – all the sounds in the language – between 120 and 130 of them – the effect of context on the sound – these are form ed into … – transform s the phonem es into allophones morphemes • syntax – structure of sentences – sm allest unit of language that has m eaning. • sem antics – m eaning of sentences Speech Recognition Problems The Phonetic Typewriter • Developed for Finnish (a phonetic language, written as it is said) • Different people speak differently: – accent, intonation, stress, idiom, volume, etc. • Trained on one speaker, will generalise to others. • The syntax of sem antically sim ilar sentences m ay vary. • A neural network is trained to cluster together sim ilar sounds, which are then labelled with the corresponding • Background noises can interfere. character. • People often “ um m m .....” and “errr.....” • When recognising speech, the sounds uttered are • Words not enough - sem antics needed as well allocated to the closest corresponding output, and the character for that output is printed. – requires intelligence to understand a sentence – requires large dictionary of minor variations to correct general – context of the utterance often has to be known mechanism – also information about the subject and speaker – noticeably poorer performance on speakers it has not been trained e.g. even if “ Errr.... I, um, don’t like this” is recognised, it is a fairly on useless piece of information on it’s own Speech Recognition: useful? The Phonetic Typewriter (ctd) a a a ah h æ æ ø ø e e e • Single user or limited vocabulary systems e.g. com puter dictation o a a h r æ l ø y y j i • Open use, lim ited vocabulary system s can o o a h r r r g g y j i work satisfactorily e.g. som e voice activated telephone system s o o m a r m n m n j i i • general user, wide vocabulary systems … … still a problem l o u h v vm n n h hj j j • Great potential, however l u v v p d d t r h hi j – when users hands are already occupied e.g. driving, manufacturing . . u v tk k p p p r k s – for users with physical disabilities . . v k pt t p t p h s s – lightweight, m obile devices 2
Speech Synthesis Speech Synthesis: useful? The generation of speech Successful in certain constrained applications when the user: Useful – is particularly motivated to overcome problems – natural and familiar way of receiving information – has few alternatives Problem s Exam ples: – similar to recognition: prosody particularly • screen readers Additional problem s – read the textual display to the user – intrusive - needs headphones, or creates noise in the utilised by visually impaired people workplace • warning signals – transient - harder to review and browse – spoken information sometimes presented to pilots whose visual and haptic skills are already fully occupied Non-Speech Sounds Non-Speech Sounds: useful? • Dual m ode displays: boings, bangs, squeaks, clicks etc. – information presented along two different sensory channels – redundant presentation of information • commonly used for warnings and alarms – resolution of ambiguity in one mode through information in another • Evidence to show they are useful • Sound good for – fewer typing m istakes with key clicks – transient information – video gam es harder without sound – background status information • Language/ culture independent, unlike speech e.g. Sound can be used as a redundant mode in the Apple Macintosh; almost any user action (file selection, window active, disk insert, search error, copy complete, etc.) can have a different sound associated with it. Auditory Icons SonicFinder for the Macintosh • Use natural sounds to represent different types of • item s and actions on the desktop have object or action associated sounds • Natural sounds have associated sem antics which can be m apped onto sim ilar m eanings in the interaction • folders have a papery noise e.g. throwing something away ~ the sound of smashing glass • m oving files – dragging sound • Problem : not all things have associated m eanings • copying – a problem … • Additional inform ation can also be presented: sound of a liquid being poured into a receptacle – muffled sounds if object is obscured or action is in the rising pitch indicates the progress of the copy background – use of stereo allows positional information to be added • big files have louder sound than sm aller ones 3
Earcons Earcons (ctd) • Synthetic sounds used to convey information • family earcons • Structured combinations of notes (motives ) sim ilar types of earcons represent sim ilar classes of action or sim ilar objects: the fam ily of “ errors” would represent actions and objects contain syntax and operating system errors • Motives com bined to provide rich inform ation – com pound earcons • Earcons easily grouped and refined due to – m ultiple m otives com bined to m ake one m ore com positional and hierarchical nature com plicated earcon • Harder to associate with the interface task since there is no natural mapping touch Handwriting recognition • haptic interaction Handwriting is another com m unication m echanism which we are used to in day-to-day life – cutaneous perception • tactile sensation; vibrations on the skin • Technology – kinesthetics – Handwriting consists of com plex strokes and spaces • movement and position; force feedback – Captured by digitising tablet • information on shape, texture, resistance, • strokes transformed to sequence of dots temperature, comparative spatial factors – large tablets available • example technologies • suitable for digitising maps and technical drawings – sm aller devices, som e incorporating thin screens to – electronic braille displays display the inform ation – force feedback devices e.g. Phantom • PDAs such as Palm Pilot • resistance, texture • tablet PCs Handwriting recognition (ctd) gesture • applications • Problem s – gestural input - e.g. “ put that there” – personal differences in letter form ation – sign language – co-articulation effects • technology – data glove • Breakthroughs: – position sensing devices e.g MI T Media Room – stroke not just bitm ap • benefits – special ‘alphabet’ – Graffeti on PalmOS – natural form of interaction - pointing • Current state: – enhance com m unication between signing and non- – usable – even without training signing users – but m any prefer keyboards! • problem s – user dependent, variable and issues of coarticulation 4
Recommend
More recommend