The road is long and long, and how does AI TV break?

Referring to the current Internet hot words, "artificial intelligence" is essential. As technology matures, artificial intelligence has been used in a variety of industries, such as artificial smart phones, smart speakers, and artificial intelligence televisions. Referring to television, this home appliance that has occupied traditional home for a long time has gradually become a new game point for TV manufacturers under the further development of artificial intelligence technology. However, even with artificial intelligence TVs, it has not been able to save the TV market from increasing momentum. In the Yikang era, omnichannel data showed that Q1 color TV retail volume in 2017 was the lowest in five years, down 12.5% ​​year-on-year. In the first half of 2017, apart from slight sales growth in February and June, the sales situation in the rest of the month was not optimistic. The offline market decreased by 4.7% year-on-year and was down 19.3% year-on-year. The reason is that in addition to the large loss of young users and the change in family entertainment habits, there is another important point that cannot be ignored. That is to say, the current artificial intelligence TV function still has some limitations, and it is not enough to attract users to stop. Not only that, when the concept of smart homes is universally implemented, artificial intelligence televisions also work with smart speakers to carry out battles for entry into the smart home. Then, in terms of function and entrance, what are the aspects of artificial intelligence television and what are the insufficiencies? The pain of early technology, AI TV is not smart enough As the color TV market is on the verge of decline, sales are not optimistic. To find new breakthroughs in the industry, Artificial intelligence television has become the focus of many manufacturers , but soon entered the homogenization period. As a TV vendor's influx into this year's enthusiasm, artificial intelligence TV is not a new topic, but when the technology is not yet mature, this topic has been "broken". It can be said that there are still some pain points that are difficult to solve in artificial intelligence TVs currently on the market. First, many TVs basically stay in the stage of simple voice manipulation, which is far from the real artificial intelligence. For example, an artificial intelligence TV Q5K released by Changhong, most of the intelligent optimization involved in the optimization of the voice function, such as far voice, voiceprint recognition and semantic understanding of the correct meaning of semantic functions, in other artificial intelligence Few features are optimized. In this way, not only does the artificial intelligence TV function appear to be monotonous, but it is also difficult to satisfy consumers' increasingly critical eyes on home appliances. Second, the voice control functions of some artificial intelligence televisions are too mechanized, and the scope of voice manipulation is too narrow. For example, the 60A730U version of the artificial intelligence television launched by TCL has only its voice function valid in the television homepage system and does not work in third-party television software. For artificial smart TVs that can add software applications by themselves, the voice control function is dispensable if the voice control range is too limited. In addition, these artificial intelligence televisions, although they all mention the concept of voice interaction, that is, televisions fully understand the needs and wishes expressed by users. But in reality, there is still a long way to go from the true goal of seamless connection. For instance, in March this year, the Beijing News reporter conducted an on-site experience with artificial intelligence TV's speech recognition function. He discovered that Changhong, an AI TV model numbered 55D3S, could not accurately identify some complicated speech input, but recommended it for users. Other irrelevant movies. Over time, users will be greatly reduced in their expectations of "artificial intelligence" and lose confidence in TV brands. Thirdly, many artificial intelligence TVs want to take the differentiated competition route. However, the smart functions introduced are too ideal and they do not touch the actual needs of users. For example, Sharp's recently released access to life services for artificial intelligence TVs, in addition to its simple function of voice manipulation of television broadcast content, has also started from the daily life needs of users, and has developed television orders for takeout, movie tickets, and voice shopping. demand. On the surface, this artificial intelligence TV has indeed embarked on a differentiated route and the functionality is indeed close to the user's daily life. But think about it, when the user needs to order a meal, it is obviously more convenient to order meals through the mobile phone than to select meals based on the voice that is not necessarily understood through the television. In addition, although TVs can "replace" mobile phones to achieve these functions, in the process of watching TV, if you want to shop, you can still use mobile phones to make purchases to meet the goal of watching TV in full screen. Therefore, TV sets that take the differentiated competition route may not have really figured out the actual needs of users, but instead unilaterally think that “user needs”. Therefore, it will take some time to get through this differentiated route. From this point of view, these artificial intelligence TVs in the market are still generally in the early stage of artificial intelligence technology, and the combination of television and artificial intelligence is far from expected. Although many artificial intelligence televisions have already achieved the functions of speech recognition, face recognition, etc., these technologies still remain at the simplest level. Among them, the proportion of artificial applications in the field of artificial intelligence such as machine learning and intelligent recognition is smaller. Therefore, artificial intelligence television technology has not been truly mature but has been forcibly pushed into the market. This is a behavior that slowly consumes user trust for the color TV market. Insufficient wisdom and professionalism to make AI TV compete at the entrance to the smart home In the Internet of Things era, the concept of smart homes was launched with the support of emerging mobile Internet and big data technologies. Wisdom family, as the name suggests, is to provide users with the most convenient conditions, so that all home operations to achieve a convenient and fast purpose, and make people more energy-efficient, healthy and so on. Although smart households are currently unable to achieve full-scale popularity, they are optimistically estimated that this will be a huge market in the future. As a result, smart homes have become the focus of many Internet giants rushing to layout, and artificial intelligence TVs, which are the entrances to smart homes, are competing with smart speakers to launch a battle for smart home entrances. However, artificial intelligence televisions have some disadvantages compared to smart and compact speakers. For example, in March of this year, Changhong launched a new Q5N artificial intelligence TV with a price of several thousands to tens of thousands of yuan, which added the voice to control the functions of other home appliances, and strived to turn it into a smart home control center. However, let's not talk about whether the functions of such artificial intelligence televisions controlling other home products are comparable to smart speakers. However, considering the actual situation of homes, replacing a smart speaker is more convenient and cheaper than replacing a television. In terms of price, the average price of a smart speaker in the market is less than 1,000 yuan, but the price of a TV is at least 1,000 yuan. There are still thousands of yuan to purchase. It is not easy to achieve synchronous replacement with other household products. In contrast, if smart homes are fully rolled out in the future, then smart speaker function reforms will be faster and cheaper, or they will become smart home control centers. In addition, there are Haier Ali and Haili Ali TV production, in addition to providing a wealth of film and television content, but also voice control air conditioning to open and set the temperature, etc., to carry out unified control of home products. However, even if the goal of manipulating household products is achieved through television, compared with smart speakers, artificial intelligence TVs are still too "amateur." As we all know, the main function of television is to play TV content, playing a major role in people's free time and entertainment needs. Therefore, if it is the control center of a smart home, then this is also a cross-border derivative function of the TV. Therefore, the technical challenges are more difficult to take into consideration when people watch TV content and control smart home options. For smart speakers, controlling smart homes is more professional, and even some smart speakers on the market can turn on the TV. From this perspective, who is more professional at a glance. From this point of view, although artificial intelligence televisions and smart speakers are both in the layout of the smart home ecosystem with TV as the hub of the household Internet of Things, but from the actual comparison of certain angles, artificial intelligence TVs are at a disadvantage. Status, no smart speakers so flexible and compact. Technological innovation and cater to demand are fundamental to the qualitative change of AI TV From the revolution in imaging technology to the revolution in internet, the television industry has experienced multiple industrial changes and has made many advancements. Until now, the television industry has entered the stage of artificial intelligence. It has to be said that this is worthy of attention and expectation, because the television industry is entering the stage of technological innovation. The National Digital Audio & Video and Multimedia Product Quality Supervision and Inspection Center has provided data on the performance of artificial intelligence TVs such as storm recognition, Changhong, micro whale, millet, etc. The average of recognition accuracy is 97%, and the accuracy of smart recommendation is 93%, which can basically meet the voice interaction requirements of ordinary users. From this point of view, artificial intelligence televisions have been continuously developed and improved, and they have become more closely integrated with humans and become more user-friendly. However, since artificial intelligence televisions are currently on the path of initial evolution, there is obviously more room for future development. First of all, the future development of artificial intelligence TVs should focus on basic voice recognition technology and other functions. After making breakthroughs, they will be pushed into the market, and will be developed in more intelligent areas such as emotion recognition, so as to truly understand users. Users achieve the ultimate goal of seamless docking. Secondly, for smart home entry battles, AI TVs should be viewed more rationally. Although this is a derivative of TV functionality, it will play an important role in the future of smart homes. Therefore, in order to compete with smart home appliances such as smart speakers, artificial smart TVs are not only better at controlling the performance of home products, but also have to make breakthroughs in updating their own control functions in order to keep up with the upgrading of smart home products. pace of. In a comprehensive view, in the step-by-step exploration of artificial intelligence televisions, vendors are trying to move closer to consumer demand. In the future, we must truly achieve the ultimate goal of “connecting with the soul” of users. I am afraid that we must go through a series of technological innovations and repeated trial and error.