.png){kind=icon}
Jingkuan Song
H指数:57
被引数:1.4万
个人简介
A full professor with Tongji University. He worked in Univesity of Electronic Science and Technology of China(UESTC) as a professor (2017-2024), Columbia University as a Postdoc Research Scientist (2016-2017), and University of Trento as a Research Fellow (2014-2016) (advised by Prof. Nicu Sebe). He obtained his PhD degree from The University of Queensland (UQ), Australia (advised by Prof. Heng Tao Shen). His research interests mainly focus on Multimedia Compact Repersentation and Analysis. He was the winner of the Best Paper Award in ICPR (2016, Mexico), Best Student Paper Award in Australian Database Conference (2017, Australia), Best Paper Honorable Mention Award in SIGIR (2017, Japan), Best Paper Runner-up Award in ApWEB (2019, China) and ACM SIGMM Rising Star Award 2021.
教育背景
暂无内容
工作经历
2017.01 – 2024.12
Univesity of Electronic Science and Technology of China(UESTC)
professor
2016.01 – 2017.12
Columbia University
Postdoc Research Scientist
2014.01 – 2016.12
University of Trento
Research Fellow
荣誉奖项
- Best Paper Award in ICPR（2016）｜ICPR
- Best Student Paper Award in Australian Database Conference（2017）｜Australian Database Conference
- Best Paper Honorable Mention Award in SIGIR（2017）｜SIGIR
- Best Paper Runner-up Award in ApWEB（2019）｜ApWEB
- ACM SIGMM Rising Star Award（2021）｜ACM SIGMM
A2VAD：属性增强的弱监督视频异常检测提示学习
Pattern Recognition
IF7.6
基于参数化上下文记忆学习的自我中心在线动作分割
对视觉语言模型的条件提示微调的深入探讨
SeMv-3D：面向通用文本到3D生成中语义与多视角一致性的并发
多智能体证据关联：可解释的仇恨视频检测
通过跨模态混合超图匹配实现通用的自我中心任务验证
从通道偏差到特征冗余：揭示少样本学习中的“少即是多”原则
基于匿名化扩散模型的隐私保护行人重识别
Pattern Recognition
IF7.6
通过消偏的信息性场景图生成
VQ-Flow：通过层次化向量量化驯服归一化流，以实现多类异常检测
通过推理不确定性引导的精炼来减轻大型视觉-语言模型的幻觉
BadCM: 针对跨模态学习的隐形后门攻击
