利用空间上下文做物件识别PPT课件.ppt

利用空间上下文寻找识别物体;Things vs. Stuff;Finding Things;Outline;Satellite Detection Example;Error Analysis;Types of Context;Types of Context;Outline;Things;Things;Stuff;Relationships;The TAS Model;Unrolled Model;Learning the Parameters;Learned Satellite Clusters;Which Relationships to Use?;Learning the Relationships;Inference;Outline;Base Detector - HOG;Results - Satellite;Results - Satellite;PASCAL VOC Challenge;Base Detector Error Analysis;Discovered Context - Bicycles;TAS Results – Bicycles;Results – VOC 2005;Results – VOC 2006;Conclusions;Merci!;Object Detection;Sliding Window Detection;Sliding Window Results;Quantitative Evaluation;Prior:Detector Only;Equations;Features: Haar wavelets;Features: Edge fragments;Histograms of oriented gradients;经常不断地学习,你就什么都知道。你知道得越多,你就越有力量 Study Constantly, And You Will Know Everything. The More You Know, The More Powerful You Will Be;谢谢你的到来 学习并没有结束，希望大家继续努力 Learning Is Not Over. I Hope You Will Continue To Work Hard 利用空间上下文寻找识别物体;Things vs. Stuff;Finding Things;Outline;Satellite Detection Example;Error Analysis;Types of Context;Types of Context;Outline;Things;Things;Stuff;Relationships;The TAS Model;Unrolled Model;Learning the Parameters;Learned Satellite Clusters;Which Relationships to Use?;Learning the Relationships;Inference;Outline;Base Detector - HOG;Results - Satellite;Results - Satellite;PASCAL VOC Challenge;Base Detector Error Analysis;Discovered Context - Bicycles;TAS Results – Bicycles;Results – VOC 2005;Results – VOC 2006;Conclusions;Merci!;Object Detection;Sliding Window Detection;Sliding Window Results;Quantitative Evaluation;Prior:Detector Only;Equations;Features: Haar wavelets;Features: Edge fragments;Histograms of oriented gradients;经常不断地学习,你就什么都知道。你知道得越多,你就越有力量 Study Constantly, And You Will Know Everything. The More You Know, The More Powerful You Will Be;谢谢你的到来 学习并没有结束，希望大家继续努力 Learning Is Not Over. I Hope You Will Continue To Work Hard