arXivDaily arXiv每日学术速递 周一至周五更新

期刊&会议

AAAI Conference on Artificial Intelligence · 会议 · Artificial Intelligence

2025-12-09 至 2025-12-09 共收录 32
2512.07760 2025-12-09 cs.CV

Modality-Aware Bias Mitigation and Invariance Learning for Unsupervised Visible-Infrared Person Re-Identification

模态感知的偏见缓解与不变性学习用于无监督的可见-红外人重识别

Menglin Wang, Xiaojin Gong, Jiachen Li, Genlin Ji

AI总结 本文提出模态感知的偏见缓解与不变性学习方法,通过改进的Jaccard距离和分割与对比策略,在无监督可见-红外人重识别中实现更可靠的跨模态关联和判别性表示学习。

Comments Accepted to AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.07218 2025-12-09 cs.CL cs.AI cs.LG

NeSTR: A Neuro-Symbolic Abductive Framework for Temporal Reasoning in Large Language Models

NeSTR:一种用于大语言模型时间推理的神经符号抽象框架

Feng Liang, Weixin Zeng, Runhao Zhao, Xiang Zhao

机构 * National Key Laboratory of Big Data and Decision(大数据与决策国家重点实验室)

AI总结 NeSTR通过结合符号表示与反思推理,提升大语言模型在复杂时间约束下的推理能力。

Comments Accepted by AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.07082 2025-12-09 cs.LG

TRACE: A Generalizable Drift Detector for Streaming Data-Driven Optimization

TRACE:一种适用于流数据驱动优化的通用漂移检测器

Yuan-Ting Zhong, Ting Huang, Xiaolin Xiao, Yue-Jiao Gong

AI总结 TRACE是一种适用于流数据驱动优化的通用漂移检测器,通过基于注意力的序列学习方法有效检测流数据中的分布变化,并展示其在不同动态环境中的适应性和有效性。

Comments Accepted by AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06977 2025-12-09 eess.IV cs.LG

Physics-Guided Diffusion Priors for Multi-Slice Reconstruction in Scientific Imaging

用于科学成像多切片重建的物理引导扩散先验

Laurentius Valdy, Richard D. Paul, Alessio Quercia, Zhuo Cao, Xuan Zhao, Hanno Scharr, Arya Bangun

AI总结 本文提出了一种结合物理约束和扩散先验的框架,用于提高多切片重建的效率和质量,适用于医学和科学成像领域。

Comments 8 pages, 5 figures, AAAI AI2ASE 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06917 2025-12-09 cs.LG

Know your Trajectory -- Trustworthy Reinforcement Learning deployment through Importance-Based Trajectory Analysis

了解你的轨迹 -- 通过基于重要性的轨迹分析实现可信的强化学习部署

Clifford F, Devika Jay, Abhishek Sarkar, Satheesh K Perepu, Santhosh G S, Kaushik Dey, Balaraman Ravindran

机构 * Clifford F(未知) Devika Jay(未知) Abhishek Sarkar(未知) Satheesh K Perepu(未知) Santhosh G S(未知) Kaushik Dey(未知) Balaraman Ravindran(未知)

AI总结 本文提出基于重要性的轨迹分析框架,通过评估轨迹级状态重要性,提升强化学习部署的可信度和可解释性。

Comments Accepted at 4th Deployable AI Workshop at AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.17397 2025-12-09 cs.CV

MCMoE: Completing Missing Modalities with Mixture of Experts for Incomplete Multimodal Action Quality Assessment

MCMoE:基于专家混合的缺失模态补全用于不完整多模态动作质量评估

Huangbiao Xu, Huanqi Wu, Xiao Ke, Junyi Wu, Rui Xu, Jinglin Xu

AI总结 MCMoE通过专家混合方法解决多模态动作质量评估中缺失模态的问题,实现单阶段训练下的多模态学习和生成。

Comments AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.11551 2025-12-09 cs.AI cs.CL

Aligning Machiavellian Agents: Behavior Steering via Test-Time Policy Shaping

对 Machiavellian 代理进行对齐:通过测试时策略塑造实现行为引导

Dena Mujtaba, Brian Hu, Anthony Hoogs, Arslan Basharat

AI总结 本文提出了一种测试时策略塑造方法,通过模型引导的策略调整,解决预训练代理在复杂环境中的伦理对齐问题,实现奖励最大化与伦理约束的平衡。

Comments Accepted to AAAI 2026 AI Alignment Track

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.08340 2025-12-09 cs.LG cs.AI

HN-MVTS: HyperNetwork-based Multivariate Time Series Forecasting

基于超网络的多变量时间序列预测

Andrey Savchenko, Oleg Kachan

AI总结 本文提出基于超网络的多变量时间序列预测模型HN-MVTS,通过生成先验和任意神经网络结合,提升预测精度和泛化能力。

Comments AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.07040 2025-12-09 cs.CV cs.CR

3D-ANC: Adaptive Neural Collapse for Robust 3D Point Cloud Recognition

3D-ANC:适应性神经坍塌用于鲁棒的3D点云识别

Yuanmin Huang, Wenxuan Li, Mi Zhang, Xiaohan Zhang, Xiaoyu You, Min Yang

AI总结 3D-ANC通过神经坍塌机制解决3D点云识别中的对抗攻击问题,结合ETF对齐和自适应训练框架提升模型鲁棒性。

Comments AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.06405 2025-12-09 cs.IR

TOOL4POI: A Tool-Augmented LLM Framework for Next POI Recommendation

TOOL4POI: 一种增强工具的LLM框架用于下一步POI推荐

Dongsheng Wang, Shen Gao, Chengrui Huang, Yuxi Huang, Ruixiang Feng, Shuo Shang

AI总结 Tool4POI通过外部检索和推理提升POI推荐性能,无需微调即可兼容现成LLM,实现对OOH场景的显著改进。

Comments A critical technical error was discovered during our internal review, leading to unreliable experimental results. The issue cannot be resolved, and the paper has also been formally withdrawn from AAAI 2026. We therefore request withdrawal of the arXiv version to maintain scientific accuracy

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.06390 2025-12-09 cs.CR cs.AI

Ghost in the Transformer: Detecting Model Reuse with Invariant Spectral Signatures

Transformer中的幽灵:通过不变频谱特征检测模型重用

Suqing Wang, Ziyang Ma, Li Xinyi, Zuchao Li

AI总结 本文提出GhostSpec,一种通过不变频谱特征验证LLM血统的方法,无需训练数据或行为修改,具有鲁棒性和高效性。

Comments Accepted at AAAI 2026 (Oral)

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.20549 2025-12-09 cs.LG cs.AI

MedGR$^2$: Breaking the Data Barrier for Medical Reasoning via Generative Reward Learning

MedGR$^2$: 通过生成奖励学习突破医学推理的数据壁垒

Weihai Zhi, Jiayan Guo, Shangyang Li

AI总结 MedGR$^2$通过生成奖励学习解决医学推理中的数据稀缺问题,实现高效训练和泛化,优于现有方法。

Comments 8 pages, 5 figures

Journal ref AAAI'2026 Main Technical Track

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.05731 2025-12-09 cs.AI cs.CL

InfiGUI-G1: Advancing GUI Grounding with Adaptive Exploration Policy Optimization

InfiGUI-G1: 通过自适应探索策略优化推进GUI接地

Yuhang Liu, Zeyu Liu, Shuanghe Zhu, Pengxiang Li, Congkai Xie, Jiasheng Wang, Xavier Hu, Xiaotian Han, Jianbo Yuan, Xinyao Wang, Shengyu Zhang, Hongxia Yang, Fei Wu

机构 * Amazon(亚马逊)

AI总结 InfiGUI-G1通过自适应探索策略优化改进GUI接地,实现显著的语义对齐和性能提升。

Comments Accepted to AAAI 2026 (Oral Presentation)

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.03929 2025-12-09 cs.AI

MOTIF: Multi-strategy Optimization via Turn-based Interactive Framework

MOTIF: 通过回合式交互框架进行多策略优化

Nguyen Viet Tuan Kiet, Dao Van Tung, Tran Cong Dao, Huynh Thi Thanh Binh

AI总结 MOTIF通过回合式交互框架实现多策略优化,提升求解器设计的自动化水平。

Comments Accepted as an oral presentation at AAAI 2026. Code available at: https://github.com/HaiAu2501/MOTIF

详情

展开后加载摘要…

URL PDF HTML 收藏
2412.10320 2025-12-09 cs.RO cs.AI

MeshA*: Efficient Path Planning With Motion Primitives

MeshA*:基于运动原语的高效路径规划

Marat Agranovskiy, Konstantin Yakovlev

AI总结 MeshA*通过在网格单元格上搜索并拟合运动原语序列,实现高效路径规划,同时保证完整性和最优性,运行时间显著优于传统方法。

Comments Accepted to AAAI-2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06811 2025-12-09 cs.CV cs.AI cs.LG cs.MM

RMAdapter: Reconstruction-based Multi-Modal Adapter for Vision-Language Models

RMAdapter: 基于重建的多模态适配器用于视觉-语言模型

Xiang Lin, Weixin Li, Shu Guo, Lihong Wang, Di Huang

AI总结 RMAdapter通过双分支架构平衡通用与任务特定知识,提升视觉-语言模型在多模态迁移学习中的性能。

Comments Accepted by AAAI 2026(Oral)

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06793 2025-12-09 cs.CV

Generalized Geometry Encoding Volume for Real-time Stereo Matching

通用几何编码体积用于实时立体匹配

Jiaxin Liu, Gangwei Xu, Xianqi Wang, Chengliang Zhang, Xin Yang

AI总结 本文提出GGEV,一种实时立体匹配网络,通过深度感知特征和动态成本聚合模块提升泛化能力,在多个基准测试中取得最佳性能。

Comments Accepted by AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06604 2025-12-09 cs.LO

Description Logics with Two Types of Definite Descriptions: Complexity, Expressiveness, and Automated Deduction

具有两种类型确定描述的描述逻辑:复杂性、表达性和自动推理

Michał Sochański, Przemysław Andrzej Wałęga, Michał Zawidzki

AI总结 本文提出两种扩展描述逻辑ALC的确定描述变体,分析其复杂性、表达性及自动推理方法,展示ALCι_G在表达性上的优势及实现的有效性。

Comments Accepted for publication at AAAI 2026; pre-print with full proofs and supplementary results

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06328 2025-12-09 cs.CV

ReCAD: Reinforcement Learning Enhanced Parametric CAD Model Generation with Vision-Language Models

ReCAD: 基于强化学习的参数化CAD模型生成增强框架

Jiahao Li, Yusheng Luo, Yunzhong Lou, Xiangdong Zhou

AI总结 ReCAD通过强化学习增强参数化CAD模型生成,利用预训练模型生成高精度CAD模型,显著提升几何精度和语义保真度。

Comments Accepted as an Oral presentation at AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06196 2025-12-09 cs.AI cs.CL

ARCANE: A Multi-Agent Framework for Interpretable and Configurable Alignment

ARCANE:一种多智能体框架,用于可解释和可配置的对齐

Charlie Masters, Marta Grześkiewicz, Stefano V. Albrecht

AI总结 ARCANE是一种多智能体框架,通过动态规则生成实现可解释和可配置的对齐,适用于复杂长期任务。

Comments Accepted to the AAAI 2026 LLAMAS Workshop (Large Language Model Agents for Multi-Agent Systems)

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06105 2025-12-09 cs.CV cs.AI

Explainable Melanoma Diagnosis with Contrastive Learning and LLM-based Report Generation

基于对比学习和大语言模型的可解释性黑色素瘤诊断

Junwen Zheng, Xinran Xu, Li Rong Wang, Chang Cai, Lucinda Siyun Tan, Dingyuan Wang, Hong Liang Tey, Xiuyi Fan

AI总结 本文提出基于对比学习和大语言模型的可解释性黑色素瘤诊断框架,通过将临床标准映射到视觉Transformer空间,实现图像与临床解释的透明连接,提升模型可解释性与临床信任度。

Comments AAAI-26-AIA

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.17008 2025-12-09 cs.LG

Mask the Redundancy: Evolving Masking Representation Learning for Multivariate Time-Series Clustering

掩盖冗余:为多变量时间序列聚类演化掩码表示学习

Zexi Tan, Xiaopeng Luo, Yunlin Liu, Yiqun Zhang

机构 * Zexi Tan, Xiaopeng Luo, Yunlin Liu, Yiqun Zhang(作者)

AI总结 本文提出EMTC方法,通过自适应掩码和多视角生成提升多变量时间序列聚类性能。

Comments Accepted to AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.13134 2025-12-09 cs.CC cs.SY eess.SY math.OC math.PR

Revealing POMDPs: Qualitative and Quantitative Analysis for Parity Objectives

揭示POMDPs:用于奇数目标的定性与定量分析

Ali Asadi, Krishnendu Chatterjee, David Lurie, Raimundo Saona

AI总结 本文研究了揭示POMDPs中奇数目标的定性与定量分析,证明其极限肯定和定量分析均为EXPTIME完全。

Comments Conference AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.12565 2025-12-09 cs.CR cs.CL

A Content-Preserving Secure Linguistic Steganography

一种内容保持的安全语言隐写术

Lingyun Xiang, Chengfu Ou, Xu He, Zhongliang Yang, Yuling Liu

AI总结 CLstega通过可控分布转换实现内容保持的安全语言隐写术,有效提升隐写通信的安全性和嵌入容量。

Comments This is the extended version of the paper accepted to AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.11865 2025-12-09 cs.GR

Learning Conjugate Direction Fields for Planar Quadrilateral Mesh Generation

学习平面四边形网格生成的共轭方向场

Jiong Tao, Yong-Liang Yang, Bailin Deng

AI总结 本文提出了一种基于神经网络的数据驱动方法,用于生成高质量的共轭方向场,以提高平面四边形网格生成的效率和质量。

Comments Accepted to AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.23796 2025-12-09 cs.AI cs.MM cs.NE

From Frustration to Fun: An Adaptive Problem-Solving Puzzle Game Powered by Genetic Algorithm

从挫败到乐趣:一种由遗传算法驱动的自适应问题解决拼图游戏

Matthew McConnell, Richard Zhao

AI总结 本文提出了一种基于遗传算法的自适应拼图游戏,通过动态调整难度提升玩家体验,探索问题解决能力的培养方法。

Comments Accepted at the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE-25)

Journal ref Proceedings of the Twenty-First AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE-25), Edmonton, Canada, November, 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.23787 2025-12-09 cs.CV cs.AI

From Unstable to Playable: Stabilizing Angry Birds Levels via Object Segmentation

从不稳定到可玩:通过物体分割稳定《愤怒的小鸟》关卡

Mahdi Farrokhimaleki, Parsa Rahmati, Richard Zhao

AI总结 通过物体分割和视觉分析,研究提出了一种方法来稳定由PCG生成的《愤怒的小鸟》关卡,提升其稳定性和可玩性。

Comments Accepted at the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE-25)

Journal ref Proceedings of the Twenty-First AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE-25), Edmonton, Canada, November, 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2504.09261 2025-12-09 cs.CV

Head-Aware KV Cache Compression for Efficient Visual Autoregressive Modeling

面向头部的KV缓存压缩以实现高效的视觉自回归建模

Ziran Qin, Youru Lv, Mingbao Lin, Hang Guo, Zeren Zhang, Danping Zou, Weiyao Lin

AI总结 HACK通过面向头部的KV缓存压缩技术,有效降低VAR模型的注意力复杂度和内存开销,实现70%的缓存压缩率和1.57倍的推理加速。

Comments Accepted by AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2503.21699 2025-12-09 cs.MM cs.AI cs.CV cs.SD eess.AS

MAVERIX: Multimodal Audio-Visual Evaluation and Recognition IndeX

MAVERIX:多模态音频视觉评估与识别指数

Liuyue Xie, Avik Kuthiala, George Z. Wei, Ce Zheng, Ananya Bal, Mosam Dabhi, Liting Wen, Taru Rustagi, Ethan Lai, Sushil Khyalia, Rohan Choudhury, Morteza Ziyadi, Xu Zhang, Hao Yang, László A. Jeni

AI总结 MAVERIX是一个用于评估多模态模型音频视觉整合能力的统一基准,通过精心设计的问题展示模型在跨模态理解上的性能,揭示了与人类水平的显著差距。

Journal ref AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2503.16929 2025-12-09 cs.CV cs.AI

TEMPLE: Incentivizing Temporal Understanding of Video Large Language Models via Progressive Pre-SFT Alignment

TEMPLE:通过渐进式预SFT对齐激励视频大语言模型的时序理解

Shicheng Li, Lei Li, Kun Ouyang, Shuhuai Ren, Yuanxin Liu, Yuanxing Zhang, Fuzheng Zhang, Lingpeng Kong, Qi Liu, Xu Sun

AI总结 TEMPLE通过渐进式预SFT对齐策略提升视频大语言模型的时序理解能力,利用直接偏好优化和课程学习增强模型对时间信息的感知。

Comments Accepted to AAAI 2026. Code available at https://github.com/lscpku/TEMPLE

详情

展开后加载摘要…

URL PDF HTML 收藏