arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

Huazhong University of Science and Technology(华中科技大学)

2026-02-10 至 2026-02-10 共收录 8
2602.08855 2026-02-10 cs.LG

Rethinking Graph Generalization through the Lens of Sharpness-Aware Minimization

通过尖锐性感知最小化重新思考图泛化

Yang Qiu, Yixiong Zou, Jun Wang

机构 * School of Computer Science and Technology, Huazhong University of Science and Technology(计算机科学与技术学院,华中科技大学)

AI总结 本文提出E2A框架,通过能量引导的潜在扰动生成伪OoD样本,提升图神经网络对分布偏移的泛化能力。

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.08524 2026-02-10 cs.CV

GeoFocus: Blending Efficient Global-to-Local Perception for Multimodal Geometry Problem-Solving

GeoFocus:融合高效全局到局部感知的多模态几何问题求解

Linger Deng, Yuliang Liu, Wenwen Yu, Zujia Zhang, Jianzhong Ju, Zhenbo Luo, Xiang Bai

机构 * School of Artificial Intelligence and Automation, Huazhong University of Science and Technology(人工智能与自动化学院,华中科技大学) MiLM Plus, Xiaomi Inc(小米公司)

AI总结 GeoFocus通过关键局部感知器和顶点语言提升多模态几何问题求解的准确性和效率

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.13524 2026-02-10 cs.CV

GO-MLVTON: Garment Occlusion-Aware Multi-Layer Virtual Try-On with Diffusion Models

GO-MLVTON: 基于服装遮挡感知的多层虚拟试穿与扩散模型

Yang Yu, Yunze Deng, Yige Zhang, Yanjie Xiao, Youkun Ou, Wenhao Hu, Mingchao Li, Bin Feng, Wenyu Liu, Dandan Zheng, Jingdong Chen

机构 * School of EIC, Huazhong University of Science and Technology, Wuhan, China(华中科技大学电子信息学院) Inclusion AI, Ant Group, Beijing, China(蚂蚁集团 inclusion AI 北京)

AI总结 GO-MLVTON通过引入服装遮挡学习模块和基于扩散模型的服装变形适配模块,实现了多层服装的高精度虚拟试穿,同时提出了MLG数据集和LACD评估指标。

Comments Accepted at ICASSP 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.13261 2026-02-10 cs.CV

Building Egocentric Procedural AI Assistant: Methods, Benchmarks, and Challenges

构建第一人称程序化AI助手:方法、基准和挑战

Junlong Li, Huaiyuan Xu, Sijie Cheng, Kejun Wu, Kim-Hui Yap, Lap-Pui Chau, Yi Wang

机构 * Department of EEE(电子工程系) The Hong Kong Polytechnic University(香港理工大学) RayNeo Tsinghua University(清华大学) Huazhong University of Science and Technology(华中科技大学) Nanyang Technological University(南洋理工大学)

AI总结 本文提出第一人称程序化AI助手,探讨其核心任务、赋能维度及挑战,通过实验评估现有方法并指明未来研究方向。

Comments Under peer-review

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.07565 2026-02-10 cs.CV

Human Identification at a Distance: Challenges, Methods and Results on the Competition HID 2025

远程人类识别:挑战、方法与HID 2025竞赛结果

Jingzhe Ma, Meng Zhang, Jianlong Yu, Kun Liu, Zunxiao Xu, Xue Cheng, Junjie Zhou, Yanfei Wang, Jiahang Li, Zepeng Wang, Kazuki Osamura, Rujie Liu, Narishige Abe, Jingjie Wang, Shunli Zhang, Haojun Xie, Jiajun Wu, Weiming Wu, Wenxiong Kang, Qingshuo Gao, Jiaming Xiong, Xianye Ben, Lei Chen, Lichen Song, Junjian Cui, Haijun Xiong, Junhao Lu, Bin Feng, Mengyuan Liu, Ji Zhou, Baoquan Zhao, Ke Xu, Yongzhen Huang, Liang Wang, Manuel J Marin-Jimenez, Md Atiqur Rahman Ahad, Shiqi Yu

机构 * Shenzhen Polytechnic University(深圳职业技术大学) Fujitsu Research and Development Center Company Ltd.(富士通研发有限公司) Beijing Jiaotong University(北京交通大学) South China University of Technology(华南理工大学) Shandong University(山东大学) EVERSPRY Wuhan University of Technology(武汉理工大学) Sun Yat-sen University(中山大学) Shanghai Jiao Tong University(上海交通大学) Fujitsu Ltd.(富士通株式会社) Huazhong University of Science and Technology(华中科技大学) Hefei University of Technology(合肥工业大学) Peking University, Shenzhen Graduate School(北京大学深圳研究生院) Beijing Normal University(北京师范大学) Watrix Technology Limited Co. Ltd.(华智科技有限公司) Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所) University of Cordoba(科尔多瓦大学) University of East London(东伦敦大学) Southern University of Science and Technology(南方科技大学)

AI总结 HID 2025竞赛评估了步态识别在远程人类识别中的挑战与进展,通过SUSTech-Competition数据集验证了算法性能的提升,达到94.2%的准确率。

Comments Accepted by IJCB 2025(https://ijcb2025.ieee-biometrics.org/competitions/)

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.18588 2026-02-10 cs.AI cs.CL cs.LG

Stability as a Liability:Systematic Breakdown of Linguistic Structure in LLMs

稳定性作为负债:在大语言模型中语言结构的系统性崩溃

Xianzhe Meng, Qiangsheng Zeng, Ling Luo, Qinghan Yang, Jiarui Hao, Wenbo Wu, Qinyu Wang, Rui Yin, Lin Qi, Renzhi Lu

机构 * School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, China(人工智能与自动化学院,华中科技大学,武汉,中国) National Institute for Data Science in Health and Medicine, Xiamen University, Xiamen, China(数据科学与医学研究院,厦门大学,厦门,中国) School of Mathematics and Statistics, Huazhong University of Science and Technology, Wuhan, China(数学与统计学院,华中科技大学,武汉,中国) School of Computer Science and Technology, Huazhong University of Science and Technology, Wuhan, China(计算机科学与技术学院,华中科技大学,武汉,中国) School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan, China(电子信息与通信学院,华中科技大学,武汉,中国) School of Electrical and Electronic Engineering, Huazhong University of Science and Technology, Wuhan, China(电气与电子工程学院,华中科技大学,武汉,中国) School of Physics, Huazhong University of Science and Technology, Wuhan, China(物理学院,华中科技大学,武汉,中国)

AI总结 本研究揭示了大语言模型中训练稳定性与生成表达性之间的不一致性,指出稳定性可能带来系统性退化,而非保证生成质量。

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.05218 2026-02-10 cs.CV

MonkeyOCR: Document Parsing with a Structure-Recognition-Relation Triplet Paradigm

MonkeyOCR: 一种基于结构-识别-关系三元组范式的文档解析方法

Zhang Li, Yuliang Liu, Qiang Liu, Zhiyin Ma, Ziyang Zhang, Shuo Zhang, Biao Yang, Zidun Guo, Jiarui Zhang, Xinyu Wang, Xiang Bai

机构 * Huazhong University of Science and Technology(华中科技大学) Kingsoft Office(金山办公)

AI总结 MonkeyOCR通过结构-识别-关系三元组范式实现文档解析,提出MonkeyDoc数据集并训练30亿参数基础模型,提升性能并优化部署效率。

详情

展开后加载摘要…

URL PDF HTML 收藏
2411.17616 2026-02-10 cs.CV

Towards Stabilized and Efficient Diffusion Transformers through Long-Skip-Connections with Spectral Constraints

通过谱约束的长跳连接实现稳定高效的扩散变换器

Guanjie Chen, Xinyu Zhao, Yucheng Zhou, Xiaoye Qu, Tianlong Chen, Yu Cheng

机构 * Shanghai Jiao Tong University(上海交通大学) Shanghai Artificial Intelligence Laboratory(上海人工智能实验室) The University of North Carolina at Chapel Hill(北卡罗来纳大学教堂山分校) SKL-IOTSC, CIS, University of Macau(澳门大学SKL-IOTSC、CIS) Huazhong University of Science and Technology(华中科技大学) The Chinese University of Hong Kong(香港中文大学)

AI总结 通过引入长跳连接和谱约束,Skip-DiT实现了图像和视频生成中的稳定高效扩散变换器,显著提升了训练和推理效率。

Comments 17 pages, 8 figures

详情

展开后加载摘要…

URL PDF HTML 收藏