Upgrade the "Shang Tang Ri Ri Xin" big model system and implement the five major product upgrades

[World Wide Web Science and Technology Report] On July 8, it was reported that at the 2023 World Artificial Intelligence Conference (WAIC), SenseTime launched a comprehensive and multi-directional upgrade of the "Shangtang SenseNova" large model system, as well as a series of large model product updates and landing achievements under the system. In addition, Shangtang also highlighted and showcased the application practices of its large model technology with various industries since its official release, including Shangtang Jueying's latest intelligent cockpit product and vehicle road cloud collaborative transportation system, as well as its implementation in production practices in industries such as finance, healthcare, e-commerce, mobile terminals, and industrial parks

[World Wide Web Science and Technology Report] On July 8, it was reported that at the 2023 World Artificial Intelligence Conference (WAIC), SenseTime launched a comprehensive and multi-directional upgrade of the "Shangtang SenseNova" large model system, as well as a series of large model product updates and landing achievements under the system. In addition, Shangtang also highlighted and showcased the application practices of its large model technology with various industries since its official release, including Shangtang Jueying's latest intelligent cockpit product and vehicle road cloud collaborative transportation system, as well as its implementation in production practices in industries such as finance, healthcare, e-commerce, mobile terminals, and industrial parks.

It is understood that Shangtang's large model system is undergoing high-speed iteration under its AGI strategic layout of "large model+large device". As a Natural language processing model with hundreds of billions of parameters, Shangtang discussed that SenseChat 2.0 version broke the limit of the input length of the big language model, and introduced model versions with different parameter levels, which can perfectly adapt to the application needs of different terminals and scenarios such as mobile terminals and cloud terminals, and reduce deployment costs. The model parameters of SenseMirage 3.0, a self-developed and generative large model of Shangtang, have increased from 1 billion since its first release in April this year to 7 billion, enabling professional photography level image detail depiction.

Not only that, the SenseAvatar 2.0 digital person generation platform of Shangtang Ruying has improved voice and mouth fluency by more than 30% compared to the 1.0 version, achieving 4K high-definition video effects, and bringing AIGC image generation and digital person singing functions. In addition, the spatial reconstruction efficiency of Shangtang Qiongyu SenseSpace 2.0 has been improved by 20%, rendering performance has been improved by 50%, and the mapping time for every 100 square kilometers of scenes can be completed in only 38 hours (with 1200TFLOPS/second computing power support); And SenseThings2.0, a Shangtang style product, achieves millimeter level precision in restoring the texture and material of small objects, and breaks through the difficulty of collecting highly reflective and mirror shaped objects.

Relying on the rapid iteration of the "Shangtang Nissin SenseNova" big model system in the underlying technology field, Shangtang is actively promoting industrial upgrading through the multimodal capability combination of big models.

Xu Li, Chairman and CEO of SenseTime, introduced: The breakthrough of big models has sparked a new round of technological revolution in artificial intelligence, followed by explosive growth in industrial demand, and new application scenarios and models are rapidly emerging. Shangtang hopes to continuously promote the leap in AI infrastructure capabilities through 'big models+big devices', not only creating more powerful basic models with universal capabilities, but also further efficiently integrating professional knowledge from different vertical fields to build more knowledgeable AI infrastructure A professional large model with more expertise can fundamentally reduce the downstream application cost and threshold of large models, allowing the industrial value of large models to bloom among thousands of industries

Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.(Email:[email protected])