Smart home voice control into a new trend

With the development of smart home industry, intelligent voice control has gradually become an industry need. According to data from the development of China's audio industry in 2014, the global intelligent voice industry has exceeded US$4.75 billion in 2014, which is nearly double the 2012 growth rate of US$2.44 billion.

Voice control is not the only way to interact with the Internet of Things in the future, but it is the most natural way. Smart home, the earliest contact with "voice" control products should be voice-activated lights. Unlike modern intelligent voice control, voice-activated lights can only be turned on when they receive a sound. After a period of quiet time, they are turned off and used in buildings. Extensive, functional not only practical but also very helpful for energy saving. Now that voice control technology is more intelligent, users can start certain functions of smart home products only through some fixed “curses”, and even more advanced products can achieve more complex “intercom” and so on. The voice instructions are used to meet the user's requirements. Voice control is what it calls a “favorite pet” for smart home products, making it easy for people to enjoy an intelligent life.

Voice control, let the product "obedient"

At present, the general home equipment control commands can basically be solved by voice. From light switches, television stations, music playback, whether it is a single product or a package, can be controlled by voice.

According to the survey, users who are willing to use voice to control homes ranked second, only lagging behind APP/computer control by 0.43%, while other centralized control methods were relatively evenly distributed. Voice control is more convenient than APP control. For example, if you use the APP control, if you want to use the APP control, you need to open the APP first, then touch the switch, and then exit the APP. Using the voice control, you only need to say “turn off the light” to make it easier and lighter. And easy to see.

There are many solutions that can now implement voice control, but the actual use of them is very different. After all, the core difficulty in voice control in smart homes is far-field recognition. Especially when controlling TVs and air conditioners, users are generally far away from the devices. Therefore, far-field speech recognition is a killer technology in the smart home environment. At present, Sibite uses the "microphone array" technology and has been able to achieve a far-field recognition rate of 5 meters to more than 92%, to meet the basic needs of family life.

Voiceprint recognition allows the product to recognize the master

Voiceprint recognition includes speaker identification and speaker verification. The former is used to determine which voice a particular segment is talking about; the latter is used to confirm whether a segment of speech is spoken by a designated person. The use of voiceprint recognition in industrial military is very extensive. In the civil sphere, especially in the smart home industry, it is mainly used in some risk control areas, including security systems, password identification, and account login.

In all biometric authentication methods, voiceprint is undoubtedly the safest and most efficient new gameplay.

Using their own proprietary voice as the identification code, so that household intelligent electronic members only listen to their own words, all device control is based on identity identification and confirmation. For example, the general “security door” uses material encryption, and some advanced ones use password settings. If voice recognition technology is added, only the owner can use a specific vocabulary to unlock. This will undoubtedly allow “security” to be further upgraded.

Human-computer interaction, allowing products to "open up"

Intelligent voice technology can do more than just voice control, but also allow homes to achieve true human-computer interaction. According to Yu Kai, chief scientist at Scipto Group, the system that uses speech to interact will be smarter and can accurately understand the user's intent to communicate through voice in a variety of environments, thereby optimizing the user's life efficiency.

As early as 2013, SBCI launched “Dialogue Workshop”, providing developers with a packaged SDK free of charge, and supporting both Apple and Android platforms synchronously. This is equivalent to installing “human brain” for smart devices and assembling “listening” and “speaking”. "Comprehensive" "dialogue" and "sensory" 5 groups of dialogue ability. Not only is it possible to understand the master's instructions, but also to answer questions exactly, and even “open up” questions when there is doubt, this is the ultimate application of voice technology in smart home devices.

Cloud or local, this is not an undeniable problem, voice interaction technology in the networking environment, will be able to play a greater role. But it doesn't mean that you can't play intelligent voice when you are offline.

Whether or not networking is required is mainly to look at the functional requirements of the equipment. Nowadays, voice technology has been completely implemented by algorithms. It is not necessary to rely on chip hardware. It is only necessary to write related programs in the hardware of the device itself. For this reason, local information storage is limited, and some complex functions, such as mass search, multiple controls, etc., must upload voice information to the cloud for processing. Simple control and identification, for example, can be completely resolved locally.

At present, smart home products are mainly used in TVs and wifi speakers, but as more and more companies enter the market, this situation may change rapidly. Sibi CEO Gao Shixing said that the core of smart home is "humanization", so that home products more in line with the needs of users, whether it is TV, speakers, or routers, lamps, as long as the "human experience" Ultimately, there is a chance to become a real explosion.