In the field of intelligent cockpit, traditional passive interaction has migrated to active interaction, and the next step will be to consider how to give full play to the advantages of artificial intelligence in more vertical fields and market segments.
General artificial intelligence will continue to improve the personalized experience of users, so that the use scenarios expand from boarding, driving, parking, and leaving the car to entertainment, office, shopping, rest, and more application scenarios can be explored.
On April 18, SenseTime's "SenseNova" large model system was unveiled at the 20th Shanghai International Automobile Industry Exhibition, which was the first time that SenseTime demonstrated the combination of its Chinese language large model "SenseChat" with the cabin.
Xu Liang, Vice President of SenseTime's Intelligent Cabin, told The Paper (www.thepaper.cn) that "SenseNova" has a variety of generative AI capabilities such as language models and digital humans, which can provide intelligent multi-modal human-vehicle interaction experience for the cabin.
The Ultimate Future Capsule is equipped with a large model of Chinese language "Talk SenseChat".
When the surging technology reporter came to the front of the Peerless Future display cabin, SenseTime AI Digital Man used facial recognition to judge the reporter as a guest after welcoming him and guided him to sit down.
Xu Liang said that if the identity information is not entered in the display cabin in advance, it will be recognized as a guest by a digital person, and if it is entered in advance, it will be recognized as a driver. In the exhibition cabin, "Discuss SenseChat" will interact with guests and recommend personalized experience content.
The Future Capsule is equipped with SenseTime's AI digital human, which can perform facial recognition on passengers.
When the surging technology reporter asked the "Discuss SenseChat" equipped with the display cabin if it could introduce the Shanghai Auto Show, the "Discussion SenseChat" quickly gave a detailed answer, and said that the Shanghai Auto Show is one of the most important exhibitions in the Chinese automotive industry. When asked if it was possible to make a trip plan for the next 5 days in Shanghai, "Discuss SenseChat" listed a number of attractions and restaurants in Shanghai, and planned the itinerary for 5 days in detail.
"Discuss SenseChat" introduces the Shanghai Auto Show.
It is reported that the display cabin can also use SenseTime's AIGC (artificial intelligence generated content) capability to automatically generate photos of various styles for guests, providing users in need with "short videos in travel" services. After children get on the bus, "Discuss SenseChat" automatically recommends entering the children's mode, playing the role of a large model of story language. The pod also features features such as health detection and an AR (augmented reality) camera. SenseChat also supports external third-party applications to help users quickly achieve goals such as meeting management, shopping, and travel.
Xu Liang said that in the field of intelligent cockpit, traditional passive interaction has migrated to active interaction, and the next step will be to consider how to give full play to the advantages of artificial intelligence in more vertical fields and market segments. "On the basis of the large model with 180 billion parameters, our main focus is how to achieve more applications in more vertical fields and market segments, so we will develop the large model more in the direction of human-computer interaction." We feel that these capabilities, once realized in the cockpit, will completely disrupt the experience of the intelligent cockpit. ”
On April 10, Xu Li, Chairman and CEO of SenseTime, released the "SenseNova" large model system at the SenseTime Technology Exchange Day, including the Chinese language large model application platform "SenseChat", the image generation model "SenseMirage", the AI digital human video generation platform "SenseAvatar", the 3D content generation platform "Qiongyu SenseSpace" and "SenseThings".
According to Xu Liang, the "Daily New SenseNova" large model system is based on the strategic layout of "big model + big computing power" to promote the development of AGI (general artificial intelligence), and SenseTime compares it to a large model "supermarket", in this supermarket, there is a trinity of data, model training and deployment of a flywheel as the basis for upstream and downstream cooperation.
Xu Liang said that in the future, general artificial intelligence will enable the base model to understand the spatial environment, perceive the user state, realize multi-modal instructions, as well as a series of capabilities such as multi-round dialogue and content generation, which can continuously improve the user's personalized experience, and also expand the use scenarios from boarding, driving, parking, and leaving the car to entertainment, office, shopping, rest, and explore more application scenarios.