Speech as User Interface and Drive for High Quality Voice Set Tone
for Consumer Electronics
Traditionally, phones have been our main means of communication and current voice processing solutions are optimized for phones being held within a few centimeters of the user’s mouth (near-field conditions). Today, people communicate through and controla broader range of connected devices, where the device will not likely be held close to the user’s mouth and in environments where the surrounding noise level may be high. In these situations, where traditional near-field voice processing solutions fail, Conexant’s AudioSmartTM far-field voice processing solution mitigates the effects of these challenging conditions and enables the clear voice communication and accurate speech control that users expect. For example, the ability to have an intelligible video chat with friends on a tablet from a noisy coffee shop, or control and communicate with a smart phone hands-free while driving despite surrounding wind and road noise. High quality far-field voice processing is not only increasingly important for current applications such as smart phones, tablets, PCs and TVs, it is a requirement for emerging applications where the traditional touch interface does not apply.
This week at the 2014 International CES,
In contrast with existing voice processing solutions that were created for the near-field phone application, Conexant’s technology was created with the most challenging far-field environment in mind, the Smart TV. Leading the way to a smarter, voice-enabled future, Conexant has successfully demonstrated the power of its far-field voice input processing solutions for Smart TVs. The recently announced CX2092x, incorporating Conexant’s third generation far-field voice processing technology, offers a highly accurate, robust way to control TVs by using voice, and has been adopted by LG for its 2014 lineup of Smart TVs. Conexant also brought the world’s first far-field class microphone ADC to market. Proprietary algorithms for far-field processing help to suppress certain surrounding noises in the environment, placing the focus on the dominant voice signal in the room and ensuring an extremely high speech recognition rate and clearer voice communications when a user is at a distance from the device (within four meters). With highly accurate far-field speech recognition performance, Conexant goes one step further by enabling natural language speech control, wherein users are not tied to a specific set of commands that the end device can understand. Its high-performance voice processing allows users to speak to their TVs in a normal fashion, using words and phrases of their choice – for a truly smart audio experience.
Conexant’s AudioSmart solutions offer the optimum mixed signal and DSP technology for high-fidelity voice and audio processing. “We recognize that voice is a natural extension of the user interface,” notes
|Key features of Conexant’s AudioSmart far-field voice processing include:|
Two microphone solution with flexible mic placement options for
High hit-rate speech recognition and voice clarity – even when the
user is up to four meters from the device
Excellent voice quality, independent of the user’s angle relative
to the device or the device orientation (i.e. portrait, landscape)
Conexant’s chip solution for Smart TVs and appliances is in production and available now. The company’s software only Windows solution for PCs is sampling. The Android software only solution for smart phones, tablets and wearables will be sampling soon. Please visit www.conexant.com for more details.
Experience Smart: Conexant at CES
At 2014 International CES, Conexant is presenting an enhanced user experience at its ‘Experience SmartTM’ suite. From voice interactive applications that go mobile – to cloud imaging for printing – to video processors that deliver quality, Conexant combines its significant IP portfolio in DSP and mixed signal technology with innovative embedded software to produce highly innovative software and silicon solutions. These solutions are designed to enrich and expand audio and imaging capabilities. At the show, Conexant will be demonstrating a range of AudioSmart solutions – including those using far-field voice processing. Please visit Conexant’s Experience Smart Facebook+page&index=4&md5=68d4922472d761665e12cbea8c9d6685" rel="nofollow">Facebook page for all of the details.