近日，多模态大模型公司生数科技宣布完成新一轮数亿元融资，本轮融资由启明创投领投，达泰资本、鸿福厚德、智谱AI、老股东BV百度风投和卓源亚洲继续跟投。生数科技旗下拥有AI绘画和视频平台PixWeaver，本轮融资资金将主要用于多模态基础大模型的迭代研发、应用产品创新及市场拓展。据悉，生数科技此前已在文生图、文生3D模型、文生视频等方面取得研究成果。在2022年9月，生数科技创始成员提出了基于Transformer的网络架构U-ViT，与Sora和Stable Diffusion3背后采用的Diffusion Transformer架构DiT在架构思路与实验路径上完全一致。本轮融资由华兴资本担任独家财务顾问。
Title: Scentech Secures Hundreds of Millions in Funding
Keywords: Scentech, Funding, AI Application
News content:
Scentech, a startup focusing on multimodal large models, has recently announced the completion of a new round of hundreds of millions of yuan in financing. The round was led by Qiming Venture Partners, with participation from Datai Capital, Hongfu Hede, Zhishu AI, and existing investors BV Baidu Ventures and Zrou Asia. The funding will primarily be used for the iterative development of multimodal foundation models, application product innovation, and market expansion. Scentech’s portfolio includes the AI painting and video platform PixWeaver. The company has previously released research findings in text-to-image, text-to-3D model, and text-to-video. In September 2022, the founding members of Scentech proposed the U-ViT architecture based on the Transformer network, which aligns with the Diffusion Transformer architecture DiT used in Sora and Stable Diffusion3 in terms of architectural thinking and experimental paths. The round was exclusively advised by China Renaissance.
【来源】https://mp.weixin.qq.com/s/SG6psot1KXzXjP2k43P_hQ
Views: 1