In 2024, AI has emerged as the hottest topic in technology, with large models and generative AI accelerating the evolution of display terminals. The rapid development of AI technology is bringing unprecedented innovation opportunities to the TV industry.

On September 29, Hisense officially launched its new line of 100-inch AI TVs at its autumn product launch event. By upgrading its ULED AI image quality platform and introducing 12 industry-first TV AI agents, Hisense once again leads the way in AI TV innovation, enabling users to enjoy ultra-high image quality in any environment or scenario. At the same time, users can access their desired content and services through simple interactions, making the TV feel more like a family member, growing closer with each use, and becoming an indispensable companion in our daily lives.

AI Image Quality Expert: Exceptional Image Quality Regardless of Content or Environment

While many TV manufacturers strive to enhance AI capabilities, only Hisense has truly transformed the TV viewing experience with its hard-core image quality technology.

Hisense's new AI TVs are equipped with its exclusive ULED AI image quality platform, which leverages the company's proprietary Xinhai AI large model-powered XinXin AI image quality chip and a vast image database. Through knowledge distillation, transfer learning, and other techniques, the platform performs AI image quality calculations and parameter matching across all dimensions that affect the user's viewing experience, including environment, content, and viewing angle. This continuously empowers the client-side expert model, doubling the image super-resolution capability and significantly improving ambient light control precision. As a result, even ordinary 2K content can be upscaled to clearer 4K quality, eliminating distractions from ambient light, and ensuring that users enjoy ultra-high-definition image quality in any viewing environment or with any content.

Industry-First 12 AI TV Agents for a Fully Upgraded Large-Screen Interaction Experience

How does the AI experience on TVs stack up? Hisense delivered a top-notch answer at its autumn product launch event. Leveraging its self-developed Xinhai AI large model, Hisense meticulously crafted 12 AI agents, including movie and TV experts and gaming assistants, to form a digital partner group for large screens. These agents endow TVs with human-like thinking and service capabilities, unleashing the full potential of large-screen applications. Tailored to different user needs, Hisense's AI agents offer personalized services such as travel guides and fitness plans. Each agent has undergone extensive “expert-level training” using vast databases, enabling them to learn autonomously, grow continuously, and be activated with a single tap. They lead the industry in language understanding, natural conversation, and image and text creation capabilities.

Powered by the Xinhai AI large model's robust dialogue comprehension capabilities, Hisense introduces the industry's first multi-intent, full-duplex dialogue mode, allowing users to complete multiple tasks with a single command and maintain continuous interaction after the initial wake-up. This exclusive voice interruption feature significantly enhances the user interaction experience. Simultaneously, the Xinhai AI large model overcomes industry challenges related to role consistency and content continuity in multi-modal generation, enabling users to create freely within specified styles. It has already launched AI-native applications such as art wallpaper generation and children's picture book creation, enriching content creation capabilities for large screens. With a child's description, the AI can generate a storybook.

For example, the “Sports Agent” can answer questions about various sports types and events, including rules, tournament formats, player information, sports culture, and history. It can explain concepts like offside in soccer, the number of events at the Olympics, and the origins of the Olympic Games, helping users better understand sports events during major competitions like the Olympics.

The “Gaming Agent” assists users by opening games with a single command, automatically setting game parameters, and quickly providing game backstories, level guides, and other information. It becomes a valuable teammate, enabling users to access guides while playing.

The “Chef Master” agent provides information on various cuisines, including their characteristics, preparation methods, and ingredients. Users can obtain precise food recommendations by asking questions like, “What are some quick meals suitable for office workers in Beijing and how do I make them?” or “What are the traditional Tibetan dishes and what do they reflect about Tibetan food culture?”

Computing power, algorithms, and data form the three technical foundations of AI TVs. Chips provide computing power, large models enable algorithms, and cloud platforms facilitate data management. As the core technology of Hisense's AI TVs, the self-developed Xinhai AI large model surpasses traditional command-based language dialogue limitations, achieving “personal assistant”-level natural dialogue interactions through cross-scenario task planning. Additionally, its multi-modal large model breaks through technical barriers in generating images from text and videos from images and text, enabling users to create freely. Currently, the Xinhai AI large model ranks first in the authoritative evaluation benchmark C-Eval across the industry.