视启未来——两大AI领军人物看中的"空间智能模型"公司

Core Viewpoint - The article discusses the advancements in "spatial intelligent models" led by the company Vision Future, highlighting its competitive edge in the AI field, particularly in visual models, and the support from prominent figures in AI research [2][5][6]. Group 1: Company Background - Vision Future was founded by Dr. Zhang Lei, a prominent figure in AI, who has developed the state-of-the-art visual model Grounding DINO 1.5, outperforming major competitors like Google and Meta [5][6]. - The company has received significant backing from renowned AI experts, including Academicians Zhang Bo and Shen Xiangyang, who serve as advisors [6][8]. Group 2: Technological Advancements - Vision Future's DINO-X model has unique "generalized perception" capabilities, leading to partnerships with major companies like China Merchants Group and Meituan Robotics [8][9]. - The company aims to integrate spatial perception models with Visual-Language-Action (VLA) frameworks to create intelligent systems that align with physical world laws [9][11]. Group 3: Research and Development - The core research direction includes upgrading 2D perception to 3D understanding, addressing key challenges in embodied intelligence [11][12]. - The OVSeg3R model, developed under Dr. Zhang's guidance, has achieved significant breakthroughs in 3D object detection and segmentation, enhancing the capabilities of embodied intelligence [12][13]. Group 4: Market Potential - The article emphasizes the dual benefits of technological iteration and industrial integration in the spatial intelligence model sector, suggesting a bright future for Vision Future as a potential unicorn in this field [14].