Speech and Voice Recognition Markets to Grow Twelvefold

Starting from a base of $600 million in 2014, the global speech and voice biometrics market will reach $7.1 billion by 2024, according to market research firm Tractica. That represents a twelvefold increase.

The firm also projects that the cumulative revenue for the 10-year period will total $19 billion. At the same time, the number of licenses for voice and speech recognition software is expected to increase from 49 million worldwide this year to 565.8 million by 2024.

Tractica differentiates between speech recognition (listening for words, with or without context) and voice recognition (identifying people by how they form sounds). Speech recognition, it says, is generally more useful for controlling devices; voice recognition is generally more useful for identifying people.

It does point out, however, that speech and voice recognition can both be used in many of the same industries and sometimes overlap in a single solution.

Speech and voice recognition are mature technologies that are finally finding a market, with abundant opportunity during the next 10 years, says Bob Lockhart, a principal analyst at Tractica.

The recent surge, he adds, is happening "all of a sudden."

Improvements in storage capabilities, advances in biometrics algorithms, and a surge in consumer acceptance of biometrics in general are among the main drivers of the technology’s adoption, the research firm stated in the report.

"In a general sense, the processing power and telecommunications now available on smartphones and tablets enable use cases that were formerly unthinkable," Lockhart says. "Ten years ago, there was no mobile device powerful enough and no wireless network robust enough to consider mobile biometrics use cases."

Lockhart expects the largest use case for voice recognition and biometrics "by far" to be in the areas of control and authentication for mobile and wearable devices.

"On the enterprise side, call center applications, such as caller identification and fraud detection, have substantial revenue forecast," Lockhart adds. "We have also forecast that later in the period, patient authentication to access their own medical records could drive this technology." Other uses could include government IT systems protection and control of automotive systems.

Another successful use case he cites is the pensioner proof-of-life system deployed by Mexico City. Each month, the city's roughly 1,900 pensioners are required to call and speak to a voice recognition system that verifies their voices against voice prints stored in a central database. Dubbed Viva Voz, the system uses Nuance Communications' VocalPassword technology.

Still, voice technology is not likely to be the only biometrics modality available. Lockhart expects voice to compete with fingerprint, facial recognition, and eventually iris imaging. Of those, fingerprints will rank "substantially larger than voice recognition or any other modality," Lockhart says. Voice recognition will make up "a substantially larger market than facial recognition and a somewhat larger market than iris recognition," he adds.

Voice, Lockhart explains, has several unique advantages. For one, speaking "is a natural action to perform, especially with a smartphone."

Additionally, voice can be integrated with any smartphone or tablet with software only, and it is perhaps the only biometric that can be performed with a very basic-feature phone.

Lockhart also credits Apple with bringing biometrics to the mainstream with TouchID, a fingerprint recognition feature it released in 2013 for unlocking iPhone and iPad devices and making purchases in Apple digital media stores. "This has been repeatedly cited as the tipping point for the biometrics market," Lockhart says.

And while he says the underlying technology has remained pretty much the same, vendors are constantly tweaking their offerings.

"Like all other biometrics markets, this one is driven by use cases," Lockhart says. "As such, vendors are flexible in their business models, tailoring software license pricing and programs to individual industry use cases and also selling software development kits to companies that wish to build their own."