[PConline evaluation]The technological revolution has promoted a new round of competition for entrance, and artificial intelligence has stood on the wave of the new era, giving out excited and urgent laughter. The fields it covers have expanded exponentially to almost all scientific and technological products and services, and it has begun to evolve into a brand-new infrastructure. It makes people crazy and awe.
What is the entrance to the era of artificial intelligence? "Front-end interaction" and "back-end intelligence". The most well-known and widely used is speech recognition in front-end interactive technology. The processing of natural speech has become the main way of human-computer interaction in the future, and it is also the biggest entrance in the early stage of artificial intelligence.
The artificial intelligence TV we are going to talk about today is the most representative product of human-machine voice interaction in living room entertainment. Because the interaction of TV is far less abundant than that of mobile phones and personal computers, voice control has become the best means to liberate the interaction of smart TV.
However, the voice interaction on smart TV is not as perfect as expected, and there are still many daily needs that cannot be realized through voice interaction. The content of the interaction is limited to the existing settings, not to mention intelligence. The breadth and depth of TV voice usage scenarios are lacking.
Until 2018, after Hisense upgraded the VIDAA smart TV system to the VIDAA AI artificial intelligence system, a voice interaction revolution broke out in smart TV. Artificial intelligence voice interaction can be so easy to use. The new VIDAA AI system has begun to enable artificial intelligence TV to truly understand the inner needs of users.
What are the pain points of voice interaction of smart TV? Speech recognition has slow response, low recognition rate and small recognition range … … Hisense VIDAA AI artificial intelligence system, through a new upgrade, will eliminate these old interactive pain points one by one.
Speech recognition goes through several stages: sound collection-speech recognition-semantic understanding-content matching-result display. The processing of each stage is related to the speed and accuracy of the whole interaction.
Sound acquisition stageHisense adopts lossless data compression algorithm, which makes the voice data smaller and easier to transmit, and reduces the packet loss rate by 30%.
Speech recognition stageHisense cooperated with Iflytek to introduce its speech recognition engine, which reached the same level in recognition accuracy and led the industry in dialect support.
Semantic understanding stageHisense’s enhanced version of the Intention Understanding Engine has achieved the latest technological innovations in semantic analysis, fuzzy understanding, knowledge map and contextual relevance, and Hisense Voice has become the most user-friendly TV in the industry.
Content matching stageRelying on more than 20 million TV users, Hisense has established the largest database system for TV application scenarios in the industry; At the same time, on the basis of the original business scope, it has expanded the contents closely related to daily life such as clothing, food, housing and transportation.
Result display stageHisense’s software system is optimized as a whole, and the speed of application startup and interface switching is increased by 25%, and the response is super fast. Moreover, Iflytek speech synthesis scheme is introduced to support 11 kinds of voice broadcast timbres.
Comprehensive experience of VIDAA AI artificial intelligence system
Just talk about not practicing fake tricks. Since there is such a strong technical support, let’s take a look at the real strength of the VIDAA AI artificial intelligence system.
By experiencing the video, we can clearly see and hear that Hisense VIDAA AI artificial intelligence system responds very quickly to the author’s various language instructions, the feedback results are accurate, and the questions that can be answered are also very comprehensive.
This constitutes the three speech recognition standards of Hisense VIDAA AI artificial intelligence system: fast, accurate and complete. Of course, this is only a solid basic skill, and there are more extremely difficult voice interactions to open your eyes.
Hisense has achieved three revolutionary changes through the upgrade of artificial intelligence: First, it has achieved a more advanced man-machine dialogue and has an exclusive interactive experience of "five rounds of continuous dialogue". Second, 33 kinds of person relationship recognition are realized to accurately understand the user’s intention. Third, 34 kinds of life scenes are responsive, covering major life scenes such as live broadcast, education, games, shopping, music, food, Chinese-English translation, booking air tickets and train tickets, booking hotels, and inquiring about scenic spots.
Characteristic function experience of VIDAA AI artificial intelligence system
Five rounds of continuous dialogue
The most difficult part of voice interaction is the cohesion of contextual content. For example, if you want to watch a movie, you have searched through European and American blockbusters, but suddenly you want to be optimistic about suspense movies in European and American blockbusters, so you say "suspense". Here, most intelligent voices begin to be circled, and the results you may get only include the keyword "suspense". If we continue to give conditions, it will become more and more chaotic.
Contextual multi-round dialogue animation demonstration
The Hisense VIDAA AI artificial intelligence system can achieve five rounds of continuous dialogue in context. For example:
The first sentence: "I want to see a movie."
The second sentence: "European and American"
The third sentence: "Don’t be horrible."
The fourth sentence: "the latest"
The fifth sentence: "Don’t be Jackie Chan’s"
Reply at this time … …
Hisense: Find the latest European and American movies X for you.
Identification of 33 Main Characters’ Relationships
Sometimes we think of watching a movie, but we often forget the exact information of the play. We only remember one of the actors, or even the characters related to this actor. How can we quickly find the play you want to watch? At this point, the role-related search becomes extremely important.
For example, ask "Sun Li’s movies", then ask "her husband’s movies" and then ask "their movies". Three problems. Most intelligent voices can only solve the first one. But the VIDAA AI artificial intelligence system can identify the relationships of up to 33 characters, such as "Who is Sun Li’s ex-girlfriend?" "Ma Yili’s husband’s movie" can handle all these complicated relationships.
34 categories of life scene services
Smart TV has a high-quality big screen, which is a visual shock that can’t be experienced on any other device, and naturally it can accommodate more content. If the interactive mode becomes more efficient, shopping, finding food and booking air tickets from TV will become extremely intuitive and calm.
Hisense VIDAA AI artificial intelligence system moved its huge life scene service to artificial intelligence TV, and provided a high-quality interactive experience on the basis of intelligent voice interaction. After using it, I may really return to the mobile phone with a big screen.
TV Taobao shopping, you can intelligently search for "the same coat in Tamia Liu" or even "Chinese cabbage" for a few dollars, and click "Buy Now" to place an order; Food search, you can directly say "nearby food", there will be food shops praised by the US Mission, and you can place an order with a sweep of your mobile phone. If you want to buy a plane ticket, just say "air ticket from a certain place to a certain place on a certain day of a certain month" and the system will automatically help you search and Ctrip will scan the code to place an order.
6 dialects, all-round coverage
In addition, query weather and stocks, and open games and applications in the system; Simple mathematical calculation, setting alarm clock, calendar reminder; Translation, ancient poetry; Smart home control, video calling and so on can be controlled by voice.
Of course, don’t be surprised if you don’t speak ordinary, because Hisense VIDAA AI artificial intelligence system supports six dialects: Cantonese, Sichuan dialect, Mandarin with dialect accent, Hunan dialect, Shanghai dialect and Minnan dialect. Is there anything else you don’t understand?
Many TV search contents only support the search of a certain business, such as searching for plants vs. zombies, and only game applications may appear, while Hisense VIDAA AI artificial intelligence system not only appears games, but also movies, education, shopping and other contents of plants vs. zombies. Because it can realize the unified search of all services and support the search of multi-service contents such as movies, education, games, applications, commodities, music and encyclopedia.
Summary:With the increasing functionality of smart TV, we urgently need a simpler way of interaction, and voice service has taken on this important task. Hisense’s newly upgraded VIDAA AI artificial intelligence system has injected fresh blood into the pursuit of user experience and product design for smart TV. It brings us an intelligent voice service experience with faster response, stronger understanding and wider coverage. This wave of Hisense VIDAA AI artificial intelligence will continue to set off the interactive revolution of smart TV!