A Q&A with Aphrodite Brinsmead, an associate analyst at Datamonitor
We live in a world of ubiquitous computing—people stay connected with more than just a PC. To put it into context, prior to the 1980s, multiple humans interacted with one computer (a mainframe). In the ’80s, the birth of the PC changed the ratio to one human to one computer. Today, with the proliferation of PCs, mobile phones, MP3 players, GPS navigation systems, ATMs, and ticket kiosks, the ratio has shifted to one human to multiple computers. User interaction with these devices is crucial and speech recognition is making inroads into the interface.
Daniel Hong: What can we expect from PC and embedded speech recognition in 2008?
Aphrodite Brinsmead: Issues surrounding accuracy, which, tied to limitations in mobile device memory and processing power, have in the past restricted the use of embedded speech. However, this is beginning to change with increased mobile computing power and greater sharing of information between the device and networks. 2008 will see improvements in the number of handheld commands controlled by voice recognition, as well as better integration of speech dictation with SMS, email, and document editing programs.
DH: What’s helping drive growth for PC and embedded speech recognition?
AB: Improvements to both speech-based transcription and embedded speech technology have led enterprises to embrace speech recognition. In particular, speech recognition has become a key technology for the healthcare industry. In the mobile market, handheld devices are becoming more powerful and desktop functionality is being extended to mobility in the form of smart phones and email devices. This means that handsets can now support a larger number of voice-activated commands, and users will be able to access mobile applications by voice control.
DH: Are there other markets where you see PC and embedded speech recognition having greater uptake?
AB: We see PC-based speech recognition emerging in education, professional services, and consumer markets for translation, transcription, dictation, and navigation. For embedded speech, we see continued uptake in consumer markets, namely mobile devices and the automotive market. Improvements to the accuracy delivered by speech recognition applications as well as better marketing by vendors should see the technology become more widely adopted in 2008. So far transcription/dictation services have been most successful in the healthcare industry, but stronger embedded mobile and GPS speech solutions should allow speech solutions to gain more uses in alternative markets.