University of California, San Diego
PulseAugur coverage of University of California, San Diego — every cluster mentioning University of California, San Diego across labs, papers, and developer communities, ranked by signal.
3 天有情绪数据
-
Brain Corp 与加州大学圣地亚哥分校合作开发机器人导航人工智能
Brain Corp 宣布与加州大学圣地亚哥分校建立新的合作伙伴关系，旨在增强机器人在挑战性环境中的导航能力。此次合作将利用人工智能和机器人技术专长，开发更先进的自主系统。目标是使机器人在不可预测和复杂的真实世界环境中更有效地运行。
-
Western US states deploy AI cameras for wildfire detection
States in the Western US are leveraging AI-powered camera networks for early wildfire detection. ALERTCalifornia, for instance, utilizes approximately 1,240 AI-enabled cameras across the state. Human oversight plays a c…
-
人工智能在自动驾驶和机器人领域从感知走向行动
研究人员正通过关注超越简单环境识别的行动和决策，来推进自动驾驶和多智能体协作的人工智能。在CVPR 2026上展示的新研究探索了可控场景生成、逼真模拟增强以及端到端驾驶对齐，以使人工智能不仅能感知，还能参与真实世界。这些努力旨在创建更强大的AI系统，使其能够在动态环境中进行复杂的决策、学习行动和协调行为。
-
司法部指控加州大学洛杉矶分校医学院在招生中非法考虑种族因素
美国司法部指控加州大学洛杉矶分校（UCLA）医学院在其招生过程中非法考虑种族因素，导致合格的白人和亚裔美国申请人被拒。这一发现基于被录取的非裔和西班牙裔学生的平均绩点（GPA）和考试成绩较低，是在最高法院2023年推翻平权行动裁决后，对大学招生实践审查升级的背景下出现的。加州大学洛杉矶分校表示其招生基于择优原则，并正在审查相关发现；而司法部若未能达成和解，可能会采取法律行动或导致学校失去联邦资金。
-
Meta acquires robotics AI startup Assured Robot Intelligence to advance humanoid ambitions
Meta Platforms has acquired the AI robotics startup Assured Robot Intelligence (ARI) to advance its development of humanoid robots. ARI specializes in AI models that enable robots to understand and adapt to human behavi…
-
Apple researchers develop StereoFoley for object-aware stereo audio generation from video
Apple researchers have developed StereoFoley, a new framework for generating stereo audio from video that is semantically aligned, temporally synchronized, and spatially accurate. The system addresses limitations in exi…
-
AI weather forecasts may miss rare, extreme events, experts warn
While AI models are rapidly improving weather forecasting accuracy and efficiency, experts express concern about their reliability in predicting unprecedented "gray swan" events. These rare but plausible extremes, exace…
-
Eugene Yan 使用图和 NLP 技术增强推荐系统
Eugene Yan 的博客文章详细介绍了构建优于基线矩阵分解模型的推荐系统的方法。该方法涉及使用自然语言处理 (NLP) 技术，特别是 word2vec，根据产品之间的关系生成产品向量表示。然后，借鉴 DeepWalk 等基于图的学习方法，通过识别相似项来使用这些产品嵌入进行推荐。