One acquisition makes Synaptics the big winner behind smart speakers

Entering the intelligent era of the Internet of Everything, the shortcomings of traditional device control methods such as machinery, remote control, and mobile phone software are becoming increasingly prominent. In order to make control more convenient, a new human-computer interaction revolution around speakers is taking place, specifically referring to the smart speaker market where domestic and foreign giants including Amazon, Google, Apple, Xiaomi, and Tmall are competing.

Amazon Echo

Their idea is to introduce voice recognition, artificial intelligence and various API interfaces to speakers, so that people can "talk" to the speakers and achieve unified control of various networked devices at home. This kind of voice interaction puts forward new requirements for smart speakers, such as better sound pickup and noise reduction.

Looking back at the evolution of technology, the earliest large-scale application of voice interaction should be on smartphones represented by the iPhone. New users of Apple devices must have been deeply impressed by the Siri speaker that "responds to every request", and smart speakers are an extension of mobile phone voice interaction, but the latter solution is not suitable for the former.

The single-microphone solution was initially adopted in smartphones. This system can provide sound signals that meet the requirements of speech recognition in low-noise, reverberation-free conditions and close to the sound source. However, this solution fails when the sound source is far away, noisy, or even with echoes. So the industry turned to microphone arrays for help, and the results seem to be as expected.

However, at a media conference on human-machine interface innovation held by Synaptics recently, Saleel Awsare, vice president and general manager of the voice and imaging department of Synaptics, said that traditional microphone arrays cannot solve all sound source problems. Before discussing his point of view, let's first understand what a microphone array is. Conceptually, it is a system that combines a certain number of microphones to sample and process the spatial characteristics of the sound field.

Amazon Echo motherboard, green is S1053 0090 V6 microphone array

Microphone arrays can achieve functions such as speech enhancement, sound source localization, reverberation removal, and sound source signal extraction, and can greatly improve functions such as identifying the sender of commands, distinguishing sound sources, and reducing noise. However, there are still many problems with common microphone arrays:

First, the current mainstream array solutions all use six to eight microphones, which brings up the issue of size. The PCBA cannot be made smaller, which limits the device's ID design and sacrifices portability;

Secondly, when the distance between the microphone and the signal source is too far, the signal-to-noise ratio of the recorded signal will become lower, which increases the difficulty of algorithm processing;

Third, when the music volume of the smart speaker is too loud, it is difficult to wake up the smart voice assistant.

The reason for this is mainly related to the way microphone arrays solve problems, Saleel Awsare emphasized. He said that current microphone arrays all use wave shaping to solve problems such as sound pickup. If a large amount of sound wave data needs to be collected, more microphones are needed, which undoubtedly limits the size of the terminal device. In addition, this traditional solution will also have misjudgments in public environments with multiple sound sources. Coupled with challenges such as cost and power consumption, the microphone array solution urgently needs to be upgraded.

Saleel Awsare told Semiconductor Industry Observer reporters at a media conference that their technology can perfectly solve the existing problems of smart speakers.

Before this year, everyone's impression of Synaptics was probably its touch and fingerprint recognition products. But after a series of acquisitions, this giant, founded more than 30 years ago and famous in the PC and smartphone markets, has become a rising star in the Internet of Things market, and voice interaction technology is its stepping stone.

Synaptics easily solves the above problems of traditional microphone arrays by relying on its nearly unique AudioSmart® technology. Saleel Awsare said that unlike other suppliers who use arrays of six or more microphones, after adopting their AudioSmart®, developers can use only two or four microphone arrays to achieve or exceed the effect of competitors' solutions with more than six microphones, solving various problems such as noise, echo and interference in the voice input of smart speakers.

In addition, due to the reduction of microphone usage in the solution, customers can make the PCBA area smaller and the product power consumption lower, which makes it the preferred solution for the current hot smart speaker suppliers. It is reported that the global shipments of AudioSmart® have exceeded 30 million, and well-known companies such as Amazon and Baidu are also their partners.

From public information, we can see that AudioSmart® is a hardware and software solution. The main hardware component is DSP. By running unique software on this DSP, it is easy to achieve the purpose of noise reduction and echo reduction, and at the same time, fewer microphones can be used to achieve the same purpose.

Blind source separation plays a very important role in this, Saleel Awsare emphasized.

The so-called blind source separation technology refers to the process of separating each source signal from the aliased signal (observation signal) when the theoretical model of the signal and the source signal cannot be accurately known.

Saleel Awsare believes that in the future, "home control" such as smart speakers will be ubiquitous and embedded in almost all devices. However, small devices that can be seen everywhere have strict requirements for solutions, and Synaptics' solutions can just solve these problems.

By Li Shoupeng, Semiconductor Industry Observer

Today is the 1434th issue of content shared by "Semiconductor Industry Observer" for you, welcome to follow.

Follow the WeChat public account Semiconductor Industry Observation , reply to the keyword in the background to get more content

Reply Popular Science , read more popular science articles about the semiconductor industry

Reply to DRAM , read more DRAM articles

Reply Lithography , read more articles related to lithography technology

Reply Intel , see more articles related to Intel Corporation

Reply Full screen , read more articles related to full screen

Reply Dual camera , read more articles about dual cameras on mobile phones

Reply SMIC , read more articles related to SMIC

Reply TSMC , read more articles related to TSMC

Reply Exhibition , see "2017 Latest Semiconductor Exhibition and Conference Calendar"

Reply Submit your article and read "How to become a member of "Semiconductor Industry Observer""

Reply Search and you can easily find other articles that interest you!

Click to read the original text and join the Moore Elite