Historical Perspectives and Technological Breakthroughs
Harry F. Tibbals in Medical Nanotechnology and Nanomedicine, 2017
There is an inseparable relationship between advances in science and technology and advances in medicine. Understanding the mechanisms of charge transfer in the process of photosynthesis and energetics of cellular metabolism led to breakthroughs in photoelectric engineering and solar power generation. The study of human speech and hearing development had impacts on computer speech synthesis and recognition. Investigation of the world based on nanotechnology generates new tools for application in medical practice, resulting in new understanding of health and disease. Many nanoscale technology and science developments were the result of biological experiments, and nanotechnology has always been readily applied to the investigation of biological phenomena.
Digital Communication and Social Media for People with Communicative and Cognitive Disabilities
Christopher M. Hayre, Dave J. Muller, Marcia J. Scherer in Everyday Technologies in Healthcare, 2019
Our research point at speech synthesis as a beneficial function. It was described as a major prerequisite for constructing and reading messages that should be available in all remote communication applications. This finding is in line with previous research (Mattsson Müller et al., 2010). Speech synthesis has gone from expensive assistive technology to a standard function in off-the-shelf devices, on webpages and in standard software and is today aiming for a more natural and personalised voice (Mills et al., 2014). Hopefully, this development can increase access to speech synthesis for those who need it.
Imitating dysphonic voice: a suitable technique to create speech stimuli for spoken language processing tasks?
Published in Logopedics Phoniatrics Vocology, 2020
Isabel S. Schiller, Angélique Remacle, Dominique Morsomme
Finally, speech synthesis could be a way to generate different voice qualities for listening tasks. Compared to the three methods presented above, speech synthesis would offer the highest control of voice parameters over time. Distinct voice characteristics could be manipulated to obtain the voice quality of interest for the listening experiment. In the context of dysphonic voice creation, speech synthesis has primarily been performed on sustained vowels [18–26] or vowel combinations [22,27]. To our knowledge, synthesis of dysphonic voice in connected speech has only been performed by Yiu and colleagues [28,29], who addressed the problem of limited naturalness of the samples [29]. To assess the effect of dysphonic voice on spoken language processing, researchers require dysphonic samples of connected speech which sound natural. It seems that speech synthesis technology cannot yet respond to that need.
Speech intelligibility of virtual humans
Published in International Journal of Audiology, 2018
Annelies Devesse, Alexander Dudek, Astrid van Wieringen, Jan Wouters
In the last decades, many research projects related to human-computer interactions, language or auditory-visual speech processing and automated dialogue systems implemented auditory-visual speech synthesis. Human-computer interactions have for example been integrated in a tutoring system aimed at enhancing science learning in elementary school (Ward et al. 2013). In this project, children who received face-to-face tutoring from a virtual human speaker developed by Ma et al. (2006), achieved larger learning gains compared to students in control classrooms. Virtual human speakers were also implemented in speech and language tutoring applications like training systems on pronunciation (Dey, Maddock and Nicolson 2010; Peng et al. 2018) or on speech perception and production for hearing impaired individuals (Massaro and Light 2004). More recently, Schreitmüller et al. (2018) demonstrated the applicability of “MASSY”, a virtual human developed by Fagel and Clemens (2004), to investigate lipreading, visual benefit and auditory-visual integration of speech in both normal hearing and hearing impaired individuals. Even telephone conversations have been extended with virtual humans to augment speech intelligibility (Siciliano et al. 2003; Salvi et al. 2009).
Message banking: Perceptions of persons with motor neuron disease, significant others and clinicians
Published in International Journal of Speech-Language Pathology, 2018
Imke Oosthuizen, Shakila Dada, Juan Bornman, Rajinder Koul
Message banking, is an intervention strategy that requires little learning and could facilitate more effective communication for persons with MND (Costello, 2014). It consists of the recording of whole messages by the person with MND in his/her own natural voice and storing these electronically for later use. Given that message banking allows for recording legacy messages in one’s own voice (Costello, 2014), its main aim is to maintain social closeness between the person with MND and his/her significant other (SO). Message banking may not be confused with “voice banking”, which refers to the process where large inventories of one’s speech sounds are recorded to create a synthetic or computer voice that sounds like the person’s own voice (Creer, Cunningham, Green, & Yamagishi, 2013; Roy et al., 2006), technology that was originally developed in the 1980s (Yamagishi, Veaux, King, & Renals, 2012). Voice banking allows users to bank their voice for a time in the future when they need to use an AAC device. A text to speech synthesis voice can be created from these banked speech samples and used in a speech generating device as a personal voice (Mills, Bunnell, & Patel, 2014). This technology is currently being widely investigated as it could allow for the development of personalised voice characteristics, and vocal identity which could possibly reduce social distance between communication partners (Yamagishi et al., 2012). The importance of voice as an identifying feature in terms of age gender, nationality, mood et centra, was recently highlighted by Nathanson (2017). Furthermore, the importance of emotional nuance in human speech was also emphasised (Pullin & Hennig, 2015).
Related Knowledge Centers
- Phonetics
- Visual Impairment
- Vocal Tract
- Symbolic Linguistic Representation
- Phone
- Diphone
- Reading Disability
- Prosody
- Vowel
- Pattern Playback