-
Grad-CAM visualization of feature attention patterns for SDA and CMDA. Highlighted regions (red/yellow) indicate areas that contribute most to themodel’s classification decision2026-09-10ACM International Conference on MultimediaHiQu3DGen: Consistent Multi-View Guided High Quality 3D Asset GenerationZhenyu Li,Shanshan Gao,Guangshun Wei✉,Jianing Zhang,Wenping Wang,Yuanfeng Zhou✉AbstractHigh quality 3D asset generation remains a long-standing challenge in computer graphics, with the core difficulty lying in accurately recovering geometry and appearance that are consistent with a given ... -
Collaborative Model and Data Adaptation at Test Time2026-06-26IEEE Transactions on Circuits and Systems for Video Technology (TCSVT)Collaborative Model and Data Adaptation at Test TimeChunyun Zhang , Fujun Yang , Chaoran Cui , Member, IEEE, Shuai Gong , Wenna Wang , Xue Lin,Yonggang Qi , Member, IEEE, and Lei Zhu , Senior Member, IEEEINTRODUCTIONDEEP learning models have achieved remarkable success in computer vision, but their performance typicallyhinges on ... -
An accuracy-efficiency balanced dehazing method via dual priors and haze condition classification2026-03-31Knowledge-Based SystemsAn accuracy-efficiency balanced dehazing method via dual priors and hazecondition classificationJing Chia,∗, Yanbing Wanga , Bingyi Youa , Shiwei WubIntroductionImage dehazing remains a challenging problem in computer visionand image processing. Its primary goal is to restore clear, detail-rich images from those degraded by atmospheric fog or haze, thereby enhancing visual quality ... -
基于拓扑先验和双分支视觉Mamba的皮肤病变分割2026-01-09计算机研究与发展基于拓扑先验和双分支视觉Mamba的皮肤病变分割耿蕾蕾，许振熙，纪则轩，武思寒，范琳伟，刘慧论文简介本文针对皮肤病变分割中普遍存在的边界模糊、拓扑结构复杂且难以保持以及伪影干扰显著等挑战，提出了一种拓扑先验引导的双分支视觉Mamba网络。该网络采用协同双编码器架构，其中拓扑分支利用多颜色空间协同的拓扑组件提取模块提取鲁棒的全局拓扑先验；图像特征分支借助高效的视觉Mamba捕捉深度视觉语义信息。... -
基于图神经网络和深度强化学习的车联网边缘计算卸载2025-12-26山东大学学报（工学）基于图神经网络和深度强化学习的车联网边缘计算卸载李明津，王倩，孟宪静，耿蕾蕾本文简介本研究研究了车联网环境下任务卸载与资源分配的优化问题，利用移动边缘计算技术，将其中一部分任务卸载到边缘服务器，在充分利用系统内的计算资源的同时，最小化任务处理的平均时延与能耗。由于车联网环境的时变性和不确定性，本研究提出了一种基于图神经网络和深度强化学习的任务卸载算法—GADA。GADA算法可以使每... -
频域纹理先验与特征增强的医学图像分割模型2025-11-11计算机科学频域纹理先验与特征增强的医学图像分割模型钟延杰，蹇木伟，张昊然，凌钰坤本文简介本文提出了一种结合频域先验与频域特征增强的创新性分割网络,旨在解决医学图像中伪装性病变区域的分割难题.通过傅里叶变换提取频域信息,并结合频域特征增强模块,模型能够更有效地捕捉图像中的全局和局部细节,从而显著提升对复杂病变区域(尤其是边缘模糊或伪装性强的病变区域)的分割能力.实验结果表明,本文方法在息肉、皮肤病变和乳腺... -
基于区块链技术构建数字信用通证“信用池”平台2023-10-23研究了基于云链融合的共享架构，设计链下（高速云）与链上（企业平行联盟链）数据协同机制；提出了基于平行联盟链的跨链可信数据交换网关，打通信息孤岛，研发了基于“信用画像”模型为高速集团内部权属单位提供信用增信服务，构建信用池机制。成果实现信用数字化、通证化，化解了集团投融资增信中的“存”、“证”和“流通”难题。成为山东高速“高金云信”的核心技术。该平台服务于集团企业三重一大监管、数字经济发展，树立...