人-物交互动作中物体和人物姿势相互上下文的建模.pptVIP

人-物交互动作中物体和人物姿势间相互上下文的建模 傅红普 模型学习框架 Hill-climbing structure learning for each activity class. foreach Iteration do - Model parameter estimation by max-margin learning; - Choose the model with the largest number of mis-classified images; - Cluster the images in the selected model into two sub-classes; - Structure learning for the two new sub-classes; end 参考文献 000 Modeling Mutual Context of Object and Human Pose in Human-Object Interaction Activities（原文） 001 Tutorial on Bayesian Networks-包含结构学习方法-加了极少中文 002 马尔科夫随机场与实时多目标跟踪算法 003 维基百科的Markov Random Field条目 004 Adaboost.ppt-汪俊（原文：Rapid Object Detection using a Boosted Cascade of Simple Features） 005 吴福朝 计算机视觉中的数学方法，P264-268，科学出版社 006 课件-SVM支持向量机-刘志丹 007 Learning to parse images of articulated bodies（原文参考文献26） * 模型推理 学习得到的模型 * 模型推理 The learned models 头部检测 躯干检测 网球拍检测 物体和身体部位的构形. Compositional Inference 合成推理 [Chen et al, 2007] * 模型推理 学习得到的模型 输出 Background and Intuition Mutual Context of Object and Human Pose Model Representation Model Learning Model Inference 试验 Conclusion Outline * * Dataset and Experiment Setup Object detection; Pose estimation; Activity classification. Tasks: [Gupta et al, 2009] Cricket defensive shot Cricket bowling Croquet shot Tennis forehand Tennis serve Volleyball smash Sport data set: 6 classes 180 training (supervised with object and part locations) 120 testing images [Gupta et al, 2009] Cricket defensive shot Cricket bowling Croquet shot Tennis forehand Tennis serve Volleyball smash Sport data set: 6 classes * Dataset and Experiment Setup Object detection; Pose estimation; Activity classification. Tasks: 180 training (supervised with object and part locations) 120 testing images Object Detection Results Cricket bat * Valid region Croquet mallet Tennis racket Volleyball Cricket ball Our Method Sliding window Pedestrian context [Andriluka et al, 2009] [Dalal Triggs, 2006] Object Detection Results * * Volleyball Cricket ball Sliding window Pedestrian context Our method Small