Junkai Yan

严俊楷 (Drinky)

我于2024年6月毕业于中山大学计算机学院，获得工学硕士学位，专业为计算机技术。我的导师是郑伟诗教授。在我攻读硕士学位之前，我也在中山大学计算机学院获得了学士学位。我很感恩在中山大学这七年的研学时光！

我的研究兴趣主要围绕计算机视觉，目标是开发通用和可推广的视觉系统。在先前几年，我主要聚焦于自监督学习 (Self-supervised Learning)、目标检测 (Object Detection) 和行人重识别 (Person Re-identification) 等相关视觉感知任务及它们在各种实际场景中的应用。如果您有任何问题想与我讨论，请随时通过以下方式与我联系：drinkyyan@gmail.com。

谷歌学术主页 / Github个人主页

News 📰

[09/26/2024] 1 篇论文被 NeurIPS 2024 接收

[07/16/2024] 1 篇论文被 ACM-MM 2024 接收为 Oral & Best Paper Nomination

[07/01/2024] 2 篇论文被 ECCV 2024 接收

[05/15/2024] 1 篇论文被 KSEM 2024 接收

[03/15/2024] 1 篇论文被 IEEE TPAMI 接收

[07/13/2023] 1 篇论文被 ICCV 2023 接收

[03/13/2023] 1 篇论文被 ICME 2023 接收为 Oral

[02/27/2023] 1 篇论文被 CVPR 2023 接收

[03/29/2022] 1 篇论文被 ICPR 2022 接收，Best Industrial Research Paper Award Candidate

Publications 📚

	Frozen-DETR: Enhancing DETR with Image Understanding from Frozen Foundation Models Shenghao Fu, Junkai Yan, Qize Yang, Xihan Wei, Xiaohua Xie, Wei-Shi Zheng NeurIPS , 2024 给您的检测器开上基础模型的外挂🚀 [ArXiv] [Code]
	Loc4Plan: Locating Before Planning for Outdoor Vision and Language Navigation Huilin Tian, Jingke Meng, Wei-Shi Zheng, Yuan-Ming Li, Junkai Yan, Yunong Zhang ACM-MM , 2024, Oral, Best Paper Nomination 让机器人像人一样学会先找到自己的定位，再看导航思考该如何走🤖 [ArXiv] [Conference Paper]
	DreamView: Injecting View-specific Text Guidance into Text-to-3D Generation *Junkai Yan,** Yipeng Gao, Qize Yang, Xihan Wei, Xuansong Xie, Ancong Wu, Wei-Shi Zheng ECCV* , 2024 使用多视角文本来精准控制您想生成的3D内容🥳 [ArXiv] [Conference Paper] [Code]
	Bridge Past and Future: Overcoming Information Asymmetry in Incremental Object Detection Qijie Mo, Shenghao Fu, Yipeng Gao, Junkai Yan, Ancong Wu, Wei-Shi Zheng ECCV , 2024 无需记忆池，帮您缓解增量目标检测中的遗忘问题🧐 [ArXiv] [Conference Paper] [Code]
	PTMA: Pre-trained Model Adaptation for Transfer Learning Xiao Li, Junkai Yan,** Jianjian Jiang, Wei-Shi Zheng, KSEM , 2024 跨域下的迁移学习不好使？PTMA来帮您🙌 [Conference Paper]
	A Versatile Framework for Multi-scene Person Re-identification Wei-Shi Zheng, Junkai Yan,** Yi-Xing Peng IEEE TPAMI , 2024 一锅端：使用一个模型就能解决各种场景下的行人重识别问题🥇 [ArXiv] [Transaction Paper] [Code]
	ASAG: Building Strong One-Decoder-Layer Sparse Detectors via Adaptive Sparse Anchor Generation Shenghao Fu, Junkai Yan, Yipeng Gao, Xiaohua Xie, Wei-Shi Zheng ICCV , 2023 您的稀疏检测器为什么非得要 6 层解码器呢🤔 [ArXiv] [Conference Paper] [Code]
	Self-supervised Cross-stage Regional Contrastive Learning for Object Detection Junkai Yan, Lingxiao Yang, Yipeng Gao, Wei-Shi Zheng ICME , 2023, Oral 提前预习下游任务的先验知识，能持续提高目标检测性能🤣 [Conference Paper] [Code]
	AsyFOD: An Asymmetric Adaptation Paradigm for Few-Shot Domain Adaptive Object Detection Yipeng Gao, Kun-Yu Lin, Junkai Yan, Yaowei Wang, Wei-Shi Zheng CVPR , 2023 跨域目标检测又缺少训练样本？看看我们的AsyFOD🪄 [Conference Paper] [Code]
	Space-correlated Contrastive Representation Learning with Multiple Instances Danming Song, Yipeng Gao, Junkai Yan, Wei Sun, Wei-Shi Zheng ICPR , 2023 缓解全局自监督对比学习过分依赖单实例先验的问题🌟 [Conference Paper] [Code]

已公开专利

联合对比损失和重建损失的自监督视觉表征学习方法（CN116310667A）

基于多重哈希的分布式存储索引方法及系统（CN114416645A）