显著性目标检测Salient Object Detection
面向 RGB、深度、热红外、水下、遥感及高分辨率场景的目标感知。Object perception across RGB, depth, thermal, underwater, remote-sensing, and high-resolution imagery.
计算机视觉 · 多模态智能Computer Vision · Multimodal Intelligence
讲师 · 计算机与信息学院(人工智能学院)Lecturer · School of Computer and Information (School of Artificial Intelligence)
Anhui Normal University · 安徽师范大学
我是安徽师范大学计算机与信息学院(人工智能学院)讲师,1996 年 3 月出生,安徽黄山人,2024 年 6 月获南开大学人工智能专业工学博士学位。研究方向包括计算机视觉与多模态智能,特别是 RGB / RGB-D / RGB-T / RGB-D-T、水下、协同、高分辨率和遥感显著性目标检测;RGB / RGB-D、植物、农作物和协同伪装目标检测;RGB-T 图像分割、高精度图像分割、多模态大语言模型用于视觉任务、医学图像分割,以及表面缺陷检测、道路提取等二值图像分割任务。在多模态领域,关注视觉语言融合在情感分析、谣言检测和讽刺检测等社交媒体任务中的应用。
I am a lecturer at the School of Computer and Information (School of Artificial Intelligence), Anhui Normal University. I was born in Huangshan, Anhui, in March 1996 and received my Ph.D. in Artificial Intelligence from Nankai University in June 2024. My research focuses on computer vision and multimodal intelligence, especially RGB / RGB-D / RGB-T / RGB-D-T, underwater, collaborative, high-resolution, and remote-sensing salient object detection; RGB / RGB-D, plant, crop, and collaborative camouflaged object detection; RGB-T image segmentation; high-precision segmentation; multimodal large language models for visual tasks; medical image segmentation; and binary segmentation tasks such as surface defect detection and road extraction. I also study vision-language fusion for social-media tasks including sentiment analysis, rumor detection, and sarcasm detection.
安徽师范大学 · 芜湖AHNU · Wuhu
论文 RA-COD 被 IEEE Transactions on Image Processing 接收。Our paper RA-COD was accepted by IEEE Transactions on Image Processing.
论文 SPP-SCL 被 AAAI 2026 接收。Our paper SPP-SCL was accepted by AAAI 2026.
指导学生在中国高校计算机大赛 AIGC 赛道华东赛区获二等奖与三等奖。Student teams received second- and third-prize awards in the East China Division of the China Collegiate Computing Contest AIGC track.
论文发表于 CVPR、ICCV、TIP 与 ACM TOMM,多项工作开放代码。New work appeared at CVPR, ICCV, TIP, and ACM TOMM, with code released for multiple projects.
围绕真实场景中的感知、分割与语义理解问题,构建高效、可靠并具有泛化能力的视觉模型。Building efficient, reliable, and generalizable vision models for perception, segmentation, and semantic understanding in real-world environments.
面向 RGB、深度、热红外、水下、遥感及高分辨率场景的目标感知。Object perception across RGB, depth, thermal, underwater, remote-sensing, and high-resolution imagery.
探索弱对比、复杂背景与跨模态条件下的伪装目标发现和分割。Discovering and segmenting concealed objects under low contrast, clutter, and multimodal conditions.
研究多模态大模型、社交媒体理解,以及息肉、超声与 PET-CT 分割。Multimodal foundation models, social-media understanding, and polyp, ultrasound, and PET-CT segmentation.