近日，2019年国际计算机视觉会议（ICCV 2019）论文录用名单发布，我院计算机视觉研究所的王振华博士、张剑华副教授、硕士研究生刘通在图模型推理方面的研究论文“New Convex Relaxations for MRF Inference with Unknown Graphs”被该会议录用。这也是我院首次作为第一完成单位在计算机视觉国际顶级会议上发表论文。该文对未知图结构的MRF推理问题进行深入研究，提出了原NP难优化问题的两种新的凸优化松弛方法，证明了所提线性松弛问题比已有松弛更加紧致，进一步导出了求解原推理问题的有效迭代算法，并成功应用于计算机视觉中的多人行为理解问题，使行为识别与理解性能有了显著提升。 该篇论文还得到澳大利亚阿德莱德大学Qinfeng Shi副教授、牛津大学M. Pawan Kumar副教授的悉心指导。

    ICCV 的全称是 IEEE International Conference on Computer Vision，即国际计算机视觉大会，被中国计算机学会（CCF）列为A类会议。 ICCV由IEEE主办，与计算机视觉模式识别会议（CVPR）和欧洲计算机视觉会议（ECCV）并称计算机视觉方向的三大顶级会议。ICCV 每两年举办一次，2019年ICCV将于10月27--11月2日在韩国首尔举办。

附：

题目：New Convex Relaxations for MRF Inference with Unknown Graphs

作者：Zhenhua Wang, Tong Liu, Qinfeng Shi, M. Pawan Kumar, Jianhua Zhang

Abstract

Treating graph structures of Markov random fields as unknown and estimating them jointly with labels have been shown to be useful for modeling human activity recognition and other related tasks. We propose two novel relaxations for solving this problem. The first is a linear programming (LP) relaxation, which is provably tighter than the existing LP relaxation. The second is a non-convex quadratic programming (QP) relaxation, which admits an efficient concave-convex procedure (CCCP). The CCCP algorithm is initialized by solving a convex QP relaxation of the problem, which is obtained by modifying the diagonal of the matrix that specifies the non-convex QP relaxation. We show that our convex QP relaxation is optimal in the sense that it minimizes the l1 norm of the diagonal modification vector. While the convex QP relaxation is not as tight as the existing and the new LP relaxations, when used in conjunction with the CCCP algorithm for the non-convex QP relaxation, it provides accurate results. We demonstrate the efficacy of our new relaxations for both synthetic data and for human activity recognition.