大场景多对象的深度社交分组网络  

Deep social grouping network for large scenes with multiple subjects

作　　者：李坤[1] 李万鹏 孙晓琨 方璐 Kun LI;Wanpeng LI;Xiaokun SUN;Lu FANG(College of Intelligence and Computing,Tianjin University,Tianjin 300350,China;Department of Electronic Engineering,Tsinghua University,Beijing 100084,China)

机构地区：[1]天津大学智能与计算学部,天津300350 [2]清华大学电子工程系,北京100084

出　　处：《中国科学：信息科学》2021年第8期1287-1301,共15页Scientia Sinica(Informationis)

基　　金：国家自然科学基金(批准号:61860206003);天津市应用基础与前沿技术研究计划(自然科学基金)(批准号:18JCYBJC19200)资助项目。

摘　　要：在计算机视觉中,群体分析越来越受到人们的关注,对图像中复杂人群进行分组是群体分析领域的基础技术需求.现有的人群社交分组方法只针对固定人数的小范围场景,不能处理真实世界中的大场景图像.本文提出首个面向十亿像素大场景图像的基于深度学习的细粒度人群社交分组框架,由一种图引导的全局到局部的划分策略与一个学习隐函数表示社交对交互模式的深度社交分组网络组成.该框架可在大范围场景图像上实现准确的人群分组.本文方法同样适用于小场景图像,在小场景图像数据集上的实验结果表明,本文提出的框架相比于现有方法取得了显著的性能提升.相关代码与训练数据即将开源.In computer vision, more attention has been paid to group analysis, and the group detection in images becomes a key technology of human analysis on groups. The existing social grouping methods only focus on small scenes with fixed number of persons and cannot deal with large scene images in the real world. This paper proposes the first fine-grained social grouping framework for gigapixel large scene images based on deep learning, which consists of a graph-guided global-to-local partition strategy and a deep grouping network that learns an implicit respresentation for social pairs. The framework has achieved accurate grouping on large scene images. Our method is also applicable to small scene images, and has outperformed the existing methods. The relevant code and the training dataset will be released soon.

关 键 词：群体 大场景图像 深度学习 社交分组 图引导 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]