Unsupervised template learning for fine-grained object recognition
来源:互联网 发布:房地产投资模型 知乎 编辑:程序博客网 时间:2024/05/24 04:45
Shulin Yang, Liefeng Bo, Jue Wang, Linda Shapiro
Washington U
Two steps:
1) aligning image regions that contain the same object part
2) extracting image features within the aligned image regions
This model is learned using an alternative algorithm, which iterates between detecting aligned image regions, and updating the template model.
dataset: Caltech-UCSD Bird200 and Stanford Dogs
Learn the template and relationships between patterns.->use LBP or other methods to extract features of each template->SVM
Template model:
Score function between templates and a give image It :
衡量三个部分:
1)fitness, 某一个模板和最为相似的图像区域的相似程度。
模板Ti在初始位置xi_l附近alpha范围内和该模板的匹配程度
对于同一image,对各模板的fitness求和
2)co-occurrence, 在同一张图像中同时出现两种模板的可能性。
3) diversity:任意两个模板之间的距离不应太短
Overall score function between templates and images
Template learning
alternating algorithm to optimize the overall score function
找出模板子集和位置
学习模板特征
学习模板间关系,权重w
总结:参数太多了,而且学习到的特征受训练集的种类影响很大,generalization ability 尚待考察。
特征层次性抽象表示?好像做不到
- Unsupervised Template Learning for Fine-Grained Object Recognition(精读)
- Unsupervised template learning for fine-grained object recognition
- Fine-Grained Crowdsourcing for Fine-Grained Recognition(精读)
- Picking Deep Filter Responses for Fine-Grained Image Recognition
- 论文提要“Hypercolumns for Object Segmentation and Fine-grained Localization”
- PS:Hypercolumns for Object Segmentation and Fine-Grained Localization__CVPR2015
- SPDA-CNN:Unifying Semantic Part Detection and Abstraction for Fine-grained Recognition
- BoxCars: 3D Boxes as CNN Input for Improved Fine-Grained Vehicle Recognition
- 论文阅读(2)--Picking Deep Filter Responses for Fine-grained Image Recognition
- 论文阅读(3)--SPDA-CNN: Unifying Semantic Part Detection and Abstraction for Fine-grained Recognition
- 噪声数据-The Unreasonable Effectiveness of Noisy Data for Fine-Grained Recognition
- SPDA-CNN:Unifying Semantic Part Detectiojn and Abstraction for Fine-grained Recognition
- 阅读小结:The Unreasonable Effectiveness of Noisy Data for Fine-Grained Recognition
- 细粒度图像识别文章 Picking Deep Filter Responses for Fine-grained Image Recognition 阅读笔记
- 论文阅读笔记 Picking Deep Filter Responses for Fine-grained Image Recognition
- 论文阅读笔记 SPDA-CNN: Unifying Semantic Part Detection and Abstraction for Fine-grained Recognition
- 车型识别“Hyper-class Augmented and Regularized Deep Learning for Fine-grained Image Classification”
- 论文阅读:Hyper-class Augmented and Regularized Deep Learning for Fine-grained Image Classification
- Android之Handler用法总结
- HDU Arc of Dream
- 9 隐含规则
- PLSA隐变量主题模型的公式推导解惑
- 一步一步教你做ios推送
- Unsupervised template learning for fine-grained object recognition
- 戏(细)说Executor框架线程池任务执行全过程(上)
- ajax内存泄漏的解决方案以及长连接和短连接的详解
- [JQuery,前端]jQuery表格插件datatables用法详解
- android 触摸(Touch)事件、点击(Click)事件的区别(详细解析)
- IOS热更新-JSPatch实现原理+Patch现场恢复
- 118. Pascal's Triangle
- adb命令.txt
- 每日记录..