Far-field speech and voice recognition smart speaker products such as Amazon Echo, Amazon Tap, Apple HomePod, and Google Home witnessed a high level of adoption within 2 years of their commercialization. With recent advancements in adaptive noise cancellation (ANC) and beamforming technology, Qualcomm, on March 2019, unevils AI enabled Qualcomm QCS400 SoC series designed to deliver advanced multi-mic far-field voice capability with highly responsive voice activation and beamforming technologies for smart speakers such as Google Home or the Amazon Echo. The launch of new chip help device manufacturers helps manufacturers accelerate the development and commercialization of smart and networked speakers.
MarketsandMarkets™ View Point:
Siddharth Rane – Assistant Manager : Semiconductor & Electronics Research, at MarketsandMarkets™, shares his Point of View as mentioned below:
Far-field speech and voice recognition is the ability to interact with a machine using natural human language from a distance ranging from 1 to 10 m. The far-field speech recognition system is used to recognize the user voice in a noisy environment based on speaker localization using a microphone array.
FIGURE 1 GLOBAL FAR-FIELD SPEECH AND VOICE RECOGNITION MARKET
Growing impact of front-end hardware components on the accuracy of speech and voice recognition and growth in voice control-based smart speakers are a few major factors driving the growth of this market. Smart speaker devices such as Apple HomePod, the Amazon Echo are using an array of multiple microphones and beamforming technology for recognitions the voice from a distance of several meters.
The improper recognition of speech and voice form the front-end systems (microphones) in noisy and harsh working environments increases the overall all word error rate (WER) in the back end of speech and voice recognition engine which is a major restraint that far-field speech recognition system manufacturing companies are facing today. However, researchers are trying to make a system using advanced front-end voice recognition technology by introducing beamforming and active noise cancellation technology, to make far-field speech and voice recognition more robust and accurate.
Role of far-field speech and voice recognition technologies in smart speaker application
Increasing need to enhance consumer convenience is fueling the demand for far-field based smart speakers, especially in smart homes. Key players in this market are Amazon.com, Inc. (Amazon) (US), Harman International (US), Apple (Apple) (US), Sonos (US), Alphabet (Alphabet) (US), Bose Corporation (Bose) (US), and Panasonic Corporation (Panasonic) (Japan). Google, Apple, and Amazon are leveraging their huge customer base and neural networks to leverage the far-field speech and voice recognition capabilities with their smart speaker products. Apple and Amazon also have a proprietary personal voice assistant integrated with their own smart speakers, which are capable of far-field speech and voice recognition capability. This reflects their strategy to leverage volume to create an ecosystem and consequently maintain market leadership in the smart speaker market by gaining a large customer base through enabling accurate and precise far-field speech and voice recognition capabilities.
Competitive scenario among far-field speech and voice recognition market providers
The presence of several major players such as Synaptics (US), Texas Instruments (US), Qualcomm (US), Sensory (US), STMicroelectronics (Switzerland), Harman (US), Andrea Electronics (US), Cirrus Logic (US), Microsemi (US), and DSP Group (US) has intensified the competition in the global far-field speech and voice recognition market. Product launches and developments, acquisitions, and partnerships have been the major strategies adopted by the leading players to grow in the market. Among all these strategies, players have widely adopted the strategies of product launches and developments to remain competitive in the market.
High accuracy of speech recognition is a vital consideration for mission-critical applications in aerospace, autonomous cars, industrial automation, and service robots. In far-field speech and voice recognition applications, front-end hardware and noise cancellation algorithm play an important role in the accuracy and effectiveness. Consumers of the speech recognition technology mainly measure its performance on the basis of recognition accuracy of the spoken words and speed. To increase the rate of consumer adoption, far-field speech recognition technology must meet the accuracy level of near-field speech recognition, which is estimated to have an error rate of 5.1% according to IBM. Moreover, the inclusion of AI assistants is expected to be flooded beyond the smart speaker to many different types of connected smart home devices and robotics application during the forecast period. All these factors drive the far-field speech and voice recognition market.