Introduction: In 2024, AI will become the biggest technology hotspot, and large models and generative AI are accelerating the evolution of display terminals. The rapid development of AI technology is bringing unprecedented innovation opportunities to the TV industry.
9month29day,Hisense officially launched a 100-inch AI new product lineup at the autumn new product launch conference,Through the newly upgradeULED AI image quality platform and the industry's first 12 major TVInters,Once again, lead the direction of AI TV innovation,Let users enjoy ultra-high image quality experience in any environment and scene。 At the same time, you can use the simplest communication to get the most desired content and services, so that TV is more and more like a family, the more you use it, the closer you are, and become an indispensable high-energy companion in our lives.
AI image quality expert, good image quality no longer picks content and environment
Although many TV companies are working hard to improve the content of AI,But only Hisense can really rely on hard core power to improve the TV picture quality experience。
Hisense's new AITV products launched this time are equipped with its uniqueULEDAIimage quality platform,Based on the new Xinxin AI image quality chip and huge image database empowered by Hisense's large model of Xinghai,Through knowledge distillation、Transfer learning, etc.,From all dimensions that affect the user's viewing environment、content、Angle of viewAIThe image quality calculation and parameter matching are carried out,Continue to empower the end-side expert model,The super resolution ability of the image is doubled,The fineness of ambient light control has been greatly improved,It can upgrade ordinary 2K pictures to clearer4K, And say goodbye to ambient light interference, so that users can enjoy ultra-clear image quality in any viewing environment and content.
The industry's first 12 AI TV intelligent twins, a comprehensive upgrade of the large-screen interactive experience
What is the AI experience like on TV? Hisense gave a high-scoring answer at this autumn new product launch conference. Hisense relies on its self-developed Xinghai AI large model,Carefully build film and television experts、Game assistants12AIItwins,Build a large screen digital partner group,Let TV have human-like thinking and service capabilities,Stimulate the full potential of large screen application scenarios。 The intelligent body released by Hisense this time is aimed at the needs of different users,Provide personalized services,Such as travel guide、Fitness plan, etc.。 Each agent has undergone "expert-level training" with a huge database, can learn independently, continue to grow, and wake up with one click, reaching the industry-leading level in language understanding, natural dialogue, image and text creation, etc.
Among them,Based on the strong dialogue understanding ability of the Xinghai large model,Hisense is the first in the industry to launch a sentence with multiple intentions、Global full-duplex dialogue mode,Realize multiple tasks to be completed in one sentence、Continuous interaction at one wake-up,And exclusive support for voice interruption,Greatly improve the user's interactive experience。 At the same time, the Xinghai large model breaks through the technical challenges of the industry such as multi-modal generation of role consistency and content continuity, realizes the free creation of users under the specified style, and has released AI native applications such as art wallpaper generation and children's picture book creation, providing rich content creation capabilities for large-screen displays. According to the child's description, it is possible to realize the AI generation of story picture books.
Among them, the "sports agent" can answer sports questions and answers about major sports types and sports events, including sports rules, competition systems, player information, sports culture and history, etc., such as how to determine offside in football, how many major events there are in the Olympic Games, and the origin of the Olympic Games, so that in some large-scale events such as the Olympic Games, it can help users better understand sports events.
"Game Agent" can help users open the game in one sentence, automatically set game parameters, quickly obtain information such as game story background and level strategy, and watch the guide while playing the game, so as to become a good teammate of "Fly" users.
The "Food Master" agent can provide information on the characteristics, practices, and ingredients of various cuisines, and users can get more accurate food recommendations by asking questions such as "What are the fast foods suitable for office workers in Beijing and how to do them", "What are the traditional Tibetan foods and what kind of food culture is reflected".
Computing power、Algorithm、Data are the three major technical bases of AITV,Chips make computing power,Large models make algorithms,Cloud platforms make data。 As HisenseAIThe core technology base of TV,Hisense's self-developed Xinghai large model breaks through the bottleneck of traditional imperative language dialogue,Through cross-scene task planning,Achieve"Real assistant"Natural dialogue interaction。 At the same time, its multi-modal large model can break through a series of technical bottlenecks such as Wensheng pictures and Wentu videos, and realize users' free creation. At present, the Xinghai model ranks first in the industry on the authoritative evaluation list C-Eval.