UK DDS oxygen sensor DceL O2 LF small-sized long-life analog signal
6 2024-03-22
Over the past year, various industries have been exploring the application of large models. The home appliance industry is no exception. At the recent AWE2024, many brands such as Hisense and Changhong showcased products that were combined with large models. The large model is showing a rapid penetration trend in the home appliance
Jiang Feng, Executive Director of the China Household Electrical Appliances Association, believes that the AI big model provides a key technological breakthrough path for the leap of household appliance intelligence level, and will also give rise to more new products and scenarios, such as intelligent butlers, home energy management, intelligent health management, intelligent chefs, intelligent caregivers, intelligent companions, etc. However, at the same time, large models also face implementation challenges in the home appliance industry.
The application of large models in the field of household appliances
Recently, Hisense‘s release of its self-developed Star Sea model has attracted attention. Li Wei, the President of Hisense Video, stated at the press conference that Hisense believes that all experiences with large screens are worth refreshing with AI. He believes that computing power, algorithms, and data are the three major technological foundations of TV AI. Chips make computing power, big models make algorithms, and cloud platforms make data.
As the core technology foundation of Hisense TV‘s AI, Xinghai Big Model is a series of independently developed models by Hisense, which use high-quality language, image and other data of tens of millions for fine training. The Xinghai Language Big Model breaks through the bottleneck of traditional imperative language dialogue and achieves a "human assistant" style natural dialogue interaction through cross scenario task planning. At the same time, its multimodal large model can break through a series of technical bottlenecks such as cultural images and cultural videos, enabling users to freely create.
According to Hisense, they have embedded the ability of large models in four scenarios: generative lifestyle service suggestions, voice driven fuzzy content retrieval, real-time English conversations, and game assistants on Hisense‘s 8k screen voice laser TV.
For example, users can directly express their desire to stay in Shanghai for three days and where they want to go to play, and help plan their itinerary. After generating this travel guide, you can scan the code and download it to your mobile phone. The content retrieval driven by large models improves the understanding ability of TV audio and video retrieval, allowing users to remember vague storylines and also use the understanding ability of large models to search and find the audio and video content they want to