
CVPR 2026现场直击:CV与机器人的物理结界被彻底打破
Quick Answer
CVPR 2026 in Denver showcases a significant shift in computer vision towards embodied intelligence, with 16,092 submissions and a 25.42% acceptance rate.
Quick Take
Chinese institutions dominate the top ten with 8 entries, led by Shanghai Jiao Tong University. Notable workshops feature Tesla and XPeng discussing foundational models for robotics, while Xiaomi wins the RoboChallenge with a 40.89% success rate.
Key Points
- CVPR 2026 received 16,092 submissions, a 24% increase from last year.
- Chinese universities secured 8 out of the top 10 accepted papers.
- Xiaomi achieved a 40.89% success rate in the RoboChallenge competition.
- Workshops featured industry leaders like Tesla and XPeng discussing robotics.
- AI Demonstrations showcased real-world applications of cutting-edge research.
Source Excerpt
6月4日,当维也纳多瑙河畔的 ICRA 2026 进入正会的最后一天,大洋彼岸的美国丹佛科罗拉多会议中心(Colorado Convention Center)已经人声鼎沸。 许多我们前几天还在维也纳见到的熟悉面孔,各大顶尖高校的学者和硬科技企业的高管,此刻都正拖着行李箱,马不停蹄地出现在丹佛的街头。 这种罕见的“全球双城赶场”现象背后,是计算机视觉(CVPR)与机器人(ICRA)两大顶会赛道正在发生的史诗级大融合。 今年的 CVPR 在丹佛当地时间 6月3日至4日率先开启 Workshop 环节,正会及重磅颁奖典礼则在 6月5日正式拉开帷幕,雷峰网已提前抵达第一现场,为您带来这份滚烫的展前观察。 011. 6万篇投稿的“诸神之战”:从感知世界,到理解物理最新官方数据显示,本届 CVPR 2026 的论文投稿量达到了惊人的 16,092 篇,较上一年激增 24%,而最终录用量在 4,090 篇左右,录用率依旧维持在残酷的 25. 42% 上下。 翻看今年的论文列表与 Workshop 议程,如果说前两年的 CVPR 还在卷图像生成和 2D 检测,那么今年丹佛的核心叙事已经彻底转向了“多模态基础模型”与“具身智能”。
在挤满人的分会场里,这种趋势尤为明显。 无论是探讨如何将视觉-语言-动作(VLA)模型部署到自动驾驶和机器人上的 WDFM-EAI(具身智能基础模型部署)Workshop,还是引入了真实机器人操控挑战赛(ManipArena Competition)的具身智能专场,都在传递一个极其明确的信号:计算机视觉已经走出了“画框识别”的屏幕舒适区,全面向具有物理法则的真实三维世界进军。 …
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from 雷峰网 AI
See more →
刚刚,GPT 5.6 发布会上,OpenAI 暴露了哪些 Agent 技术路线?
OpenAI's GPT 5.6 integrates ChatGPT and Codex, introducing a for complex task execution, with models Soul, Terra, and Luna for efficient workflow management. The release emphasizes task orchestration, contextual understanding, and robust security measures for enterprise applications.

