What are the applications of speech recognition technology in smart homes?

Understanding speech recognition technology

Automatic Speech Recognition technology is the key technology of human-machine interface in information technology. It is an advanced technology that allows the machine to convert speech signals into corresponding texts or commands by identifying and understanding the process.

Before the invention of the computer, the idea of ​​automatic speech recognition had been put on the agenda. Early vocoders can be seen as the prototype of speech recognition and synthesis. The Radio Rex toy dog ​​produced in the 1920s may be the earliest speech recognizer. When the dog's name is called, it can pop up from the base. The Bell Lab's Audry system in the 1950s was the first speech recognition system to recognize ten English numbers.

On the one hand, the development of computer technology provides the necessary hardware and software foundations for speech recognition. On the other hand, artificial neural networks were introduced into speech recognition in the 1960s. The two major breakthroughs in this era were Linear Predictive. Coding, LPC) and Dynamic Time Warp technology, these two major achievements have effectively solved the problem of feature extraction and inequality matching of speech signals. Since then, speech recognition technology has been studied by researchers as an important topic and has made substantial progress.

The most significant breakthrough in the development of speech recognition technology is the application of the Hidden Markov Model. Relevant mathematic reasoning was proposed by Baum. After Labiner et al.'s research, Li Kaifu from Carnegie Mellon University finally realized Sphinx, the first large-vocabulary speech recognition system based on Hidden Markov Model, and then the speech recognition technology. Has not been out of the HMM framework. Taking intelligent speech recognition as an example, you need to pay attention to the following issues.

Machines and computers recognize and understand natural language. To do this, we must first break down the continuous speech into words, phonemes and other units. Second, we must establish a rule to understand the semantics.

The amount of voice information is large, and it is difficult to recognize a large range of speech. The transcript mode is not only different for different speakers but also different for the same speaker. For example, a speaker's voice information is different when he speaks casually and speaks seriously. The way a person speaks will change with time. Putonghua differs from the dialects of various localities. The parties and words in different places are different, which impairs speech recognition objectively and requires the most extensive Understanding various types of speech information, therefore, the development of intelligent speech recognition is still a long way to go.

Voice ambiguity also confuses recognition. Different words may be similar when the speaker is speaking. We call them homophones. It is necessary to determine the speaker's meaning through the context of the speech to determine these words. This is not only common in Chinese, but also in English. Very common.

The phonetic characteristics of individual letters or words and words are affected by the context. Words in different contexts show different accents, pitches, volume, and voicing speeds, which is also an important aspect that hinders the correct recognition of sentence meanings by intelligent speech.

Environmental noise and interference also have a particularly serious impact on speech recognition, which, to a large extent, reduces speech recognition rates.

Voice Recognition Leads Smart Home Sound Control Tide

In the past 20 years, significant progress has been made in speech recognition technology. It has begun to move from the laboratory to the market and has spawned an intelligent voice industry. The so-called intelligent voice industry mainly refers to the industries that provide users with various services through voice synthesis technology and voice recognition technology. In general, the user only needs to issue a command to the service terminal in a spoken manner, and the corresponding service can be obtained. This industry first appeared in the 1960s, but due to the fact that the technology is not mature and the market application is limited, consumers have low awareness of it. Under the wave of the global Internet of Things, the intelligent voice industry has become a new industry in the IT field, and user awareness and market size are gradually expanding. It has to be said that the intelligent voice industry is forming a trend. This is a new blue ocean to be developed for the developing smart home industry, which is worthy of attention.

Looking at the world, the boom in voice voice driven by the intelligent voice industry has become overwhelming. In foreign countries, the world's three major IT giants (Apple, Google, Microsoft) are actively laying out their own intelligent voice market. Apple's Siri voice control function is considered a new round of intelligent voice technology revolution.

Voice control has long been used in smart home applications. Not long ago, a U.S. audiovisual company controlled Crestron automated control systems through Siri. Users could use voice to turn on lights, adjust light levels, start home theaters, and control air conditioning. Switching video channels, etc., has made the home automation function a big step forward.

In China, the establishment of the China Voice Industry Alliance has strongly promoted the acceleration of the development of upstream and downstream companies in the Chinese voice industry chain. Nowadays, the domestic intelligent voice market has already exhibited a prosperous image, with the most outstanding performance and the fiercest competition being in the smart TV field. Haier, Changhong, TCL, Hisense and other major traditional companies have embarked on the road to change and launched their own smart TVs. At the 122nd Canton Fair, Haier Air Conditioning displayed the industry's first "voice remote control", which brought convenience to users to achieve "intelligent operation."

Users only need to speak to a small voice remote controller. This voice remote controller can perform intelligent identification and give instructions to the air conditioner to achieve automatic switching, temperature adjustment, and power saving, self-cleaning, air exchange, and formaldehyde removal. Class special mode, very easy to follow. This not only greatly enhances the intelligent degree of human-machine interaction in the product, but also hopes to stimulate the pace of innovation in the home appliance industry's new intelligent technology, making it a new engine for industry growth. In addition, the communications industry has not forgotten to share a slice of this intelligent voice business. China Mobile has said that it will gradually increase its intelligent voice capabilities in customer systems, voice portals, search portals, smart homes, SMS, and e-mail. Including intelligent speech translation, this will surely help boost China Mobile's business and services to a new level.

Voice control is the most natural and convenient control method. Although the industry still faces some technical bottlenecks, we believe that the problems of accuracy and stability plaguing its applications are also expected to be solved. For smart homes, the application of voice control technology to smart homes will become one of the new trends in the future market development.