arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

ETH Zurich(苏黎世联邦理工学院)

2026-04-16 至 2026-04-16 共收录 8
2604.14025 2026-04-16 cs.CV cs.AI cs.GR

Feed-Forward 3D Scene Modeling: A Problem-Driven Perspective

前馈3D场景建模:以问题为导向的视角

Weijie Wang, Qihang Cao, Sensen Gao, Donny Y. Chen, Haofei Xu, Wenjing Bian, Songyou Peng, Tat-Jen Cham, Chuanxia Zheng, Andreas Geiger, Jianfei Cai, Jia-Wang Bian, Bohan Zhuang

机构 * Zhejiang University(浙江大学) Nanyang Technological University(南洋理工大学) Monash University(墨尔本大学) ETH Zurich(苏黎世联邦理工学院) University of Tübingen(图宾根大学)

AI总结 本文探讨了前馈3D重建的通用方法,提出以模型设计为核心的分类体系,涵盖特征增强、几何意识、模型效率等五个核心问题,旨在推动3D场景建模的标准化与可扩展性。

Comments 67 pages, 395 references. Project page: https://ff3d-survey.github.io. Code: https://github.com/ziplab/Awesome-Feed-Forward-3D. This work has been submitted to Springer for possible publication

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.14021 2026-04-16 cs.RO

Neuromorphic Spiking Ring Attractor for Proprioceptive Joint-State Estimation

类脑脉冲环状吸引子用于本体感觉关节状态估计

Federica Ferrari, Flavia Davidhi, Bernard Maacaron, Alberto Motta, Luuk van Keeken, Elisa Donati, Giacomo Indiveri, Chiara De Luca, Chiara Bartolozzi

机构 * 1 Istituto Italiano di Tecnologia, Genoa, Italy, 2 Department of Neurosurgery, University Hospital Zürich, Switzerland 3 Institute of Neuroinformatics, Univ. of Zurich ETH Zurich, Switzerland 4 DIEEI, University of Catania Catania, Italy, 5 IDLAB, University of Antwerp AI \& Algorithms, imec, Leuven, Belgium 6 Digital Society Initiative, University of Zurich, Switzerland these authors contributed equally

AI总结 本文提出一种类脑脉冲环状吸引子网络,用于在资源受限条件下实现机器人关节角度的本体感觉估计,通过自维持的群体活动实现稳定活动斑块,具有减少漂移和提高精度的优势。

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.13263 2026-04-16 cs.LG

Binomial Gradient-Based Meta-Learning for Enhanced Meta-Gradient Estimation

二项式梯度基于元学习用于增强元梯度估计

Yilang Zhang, Abraham Jaeger Mountain, Bingcong Li, Georgios B. Giannakis

机构 * Department of Electrical and Computer Engineering(电气与计算机工程系) University of Minnesota(明尼苏达大学) Department of Computer Science(计算机科学系) ETH Zürich(苏黎世联邦理工学院)

AI总结 本文提出BinomGBML方法,通过二项式展开提升元梯度估计精度,理论分析显示其误差界优于现有方法且在温和条件下呈超指数衰减,实验验证了方法的有效性。

Comments Accepted as poster at ICLR 2026. Code available at https://github.com/AbrahamJJM/binomgbml

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.11064 2026-04-16 cs.LG cs.CV

A Faster Path to Continual Learning

连续学习的更快路径

Wei Li, Hangjie Yuan, Zixiang Zhao, Borui Kang, Ziwei Liu, Tao Feng

机构 * College of Computer Science, Sichuan University, China(四川大学计算机学院) College of Computer Science and Technology, Zhejiang University, China(浙江大学计算机科学与技术学院) Photogrammetry and Remote Sensing Lab, ETH Zürich, Switzerland(苏黎世联邦理工学院摄影测量与遥感实验室) School of Computer Science, Nanjing University, China(南京大学计算机科学系) College of Computing and Data Science, Nanyang Technological University, Singapore(南洋理工大学计算与数据科学学院) Department of Computer Science and Technology, Tsinghua University, China(清华大学计算机科学与技术系)

AI总结 本文提出C-Flat Turbo,通过优化梯度计算减少连续学习的训练成本,提升效率并保持性能。

Comments Update Author Affiliations

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.29159 2026-04-16 cs.CL cs.CY cs.HC

Kwame 2.0: Human-in-the-Loop Generative AI Teaching Assistant for Large Scale Online Coding Education in Africa

Kwame 2.0:面向非洲大规模在线编程教育的人机协同生成式AI助教

George Boateng, Samuel Boateng, Victor Kumbol

机构 * ETH Zurich, Switzerland(苏黎世联邦理工学院,瑞士) Charité - Universitätsmedizin Berlin, Germany(柏林夏里特医学院,德国) Kwame AI Inc., U.S.(Kwame人工智能公司,美国)

AI总结 Kwame 2.0通过检索增强生成技术,在非洲35个国家的15个班级中为3717名学生提供及时准确的学习支持,结合人类监督和社区参与,有效提升编程课程的教学质量。

Comments 8 pages, Accepted at the 27th International Conference on Artificial Intelligence in Education (AIED 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.20481 2026-04-16 q-bio.NC cs.CL

Coherence in the brain unfolds across separable temporal regimes

大脑中的相干性在可分离的时间 regime 中展开

Davide Staub, Finn Rabe, Akhil Misra, Yves Pauli, Roya Hüppi, Ni Yang, Nils Lang, Lars Michels, Victoria Edkins, Sascha Frühholz, Iris Sommer, Wolfram Hinzen, Philipp Homan

机构 * Department of Adult Psychiatry and Psychotherapy, University of Zurich, Zurich, Switzerland(大学精神病学与心理学系,苏黎世大学,苏黎世,瑞士) Department of Translation and Language Sciences, University Pompeu Fabra, Barcelona, Spain(翻译与语言科学系,庞培法拉大学,巴塞罗那,西班牙) Department of Neuroradiology, Clinical Neuroscience Center, University Hospital Zurich, Zurich, Switzerland(神经放射学系,临床神经科学中心,苏黎世大学医院,苏黎世,瑞士) Department of Psychology, University of Oslo, Oslo, Norway(心理学系,奥斯陆大学,奥斯陆,挪威) Cognitive and Affective Neuroscience Unit, University of Zurich, Zurich, Switzerland(认知与情感神经科学单元,苏黎世大学,苏黎世,瑞士) Center for Clinical Neuroscience and Cognition, University of Groningen, Groningen, Netherlands(临床神经科学与认知中心,格罗宁根大学,格罗宁根,荷兰) Neuroscience Center Zurich, University of Zurich and ETH Zurich, Zurich, Switzerland(苏黎世神经科学中心,苏黎世大学和ETH苏黎世分校,苏黎世,瑞士)

AI总结 研究通过无标注的漂移和移位信号揭示了语言理解中相干性的神经机制,发现默认模式网络枢纽和初级听觉皮层存在不同的神经偏好。

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.10877 2026-04-16 cs.LG

Guided Transfer Learning for Discrete Diffusion Models

指导式迁移学习用于离散扩散模型

Julian Kleutgens, Claudio Battiloro, Lingkai Kong, Benjamin Grewe, Francesca Dominici, Mauricio Tec

机构 * Harvard University(哈佛大学) ETH Zürich(苏黎世联邦理工学院)

AI总结 本文提出GTL方法,通过比率引导实现离散扩散模型的迁移学习,解决小数据下模型性能问题,展示在序列数据中效果显著,但存在源分布与目标分布重叠不足时的失效模式。

Comments Accepted at ICLR 2026 ReALM-GEN Workshop 9 pages (main text) + appendix

Journal ref ICLR 2026 ReALM-GEN Workshop

详情

展开后加载摘要…

URL PDF HTML 收藏
2310.02540 2026-04-16 cs.LG cs.AI cs.DB cs.IR

Auto-FP: An Experimental Study of Automated Feature Preprocessing for Tabular Data

Auto-FP:关于表格数据自动化特征预处理的实验研究

Danrui Qi, Jinglin Peng, Yongjun He, Jiannan Wang

机构 * Simon Fraser University(西蒙弗雷泽大学) ETH Zürich(苏黎世联邦理工学院)

AI总结 本文研究了表格数据自动化特征预处理(Auto-FP)的方法,通过评估15种算法发现进化算法表现最佳,随机搜索也表现优异,分析了算法瓶颈并探讨了其扩展方向。

详情

展开后加载摘要…

URL PDF HTML 收藏