DenseBox: Unifying Landmark Localization with End to End Object Detection
来源：互联网 发布：mac snow ball 编辑：程序博客网 时间：2024/05/16 15:52
百度深度学习研究院 的目标检测工作 类似 Faster R-CNN
输入多尺度图像，经过CNN处理，输出目标框
Model Design
模型是从 19层的VGG模型演变来的。
Multi-Level Feature Fusion： 将不同卷积层的特征联系起来可以提高检测效果
Multi-Task Training： 两个任务：1）目标有无，2）目标的位置
这里使用了 landmark 来改善定位模型
KITTI 数据库 车辆检测上面的结果：
还有一个 DenseBox2 敬请期待！
0 0
- DenseBox: Unifying Landmark Localization with End to End Object Detection
- 论文笔记（1）DenseBox: Unifying Landmark Localization with End to End Object Detection
- 车牌识别“Towards End-to-End Car License Plates Detection and Recognition with Deep Neural Networks”
- End-to-end people detection in crowded scenes
- READING NOTE: Face Detection with End-to-End Integration of a ConvNet and a 3D Model
- 车牌检测识别--Towards End-to-End Car License Plates Detection and Recognition with Deep Neural Networks
- 人脸检测--Face Detection with End-to-End Integration of a ConvNet and a 3D Model
- Localization and Object Detection
- end-to-end 理解
- 语义分割--End-to-End Instance Segmentation with Recurrent Attention
- [EMNLP2017]End-to-End Neural Relation Extraction with Global Optimization
- End-to-end Concept Word Detection for Video Captioning, Retrieval, and Question Answering
- 论文阅读：《Associative Embedding:End-to-End Learning for Joint Detection and Grouping》
- 姿态检测整理--06-Associative Embedding: End-to-End Learning for Joint Detection and Grouping
- End-to-End Workload Analysis
- End-To-End Memory Networks
- End-To-End Memory Networks
- End-to-End Memory Network
- Count and Say
- android:versionCode和android:versionName 用途
- Spring MVC 3.0.5+Spring 3.0.5+MyBatis3.0.4全注解实例详解（二）
- 敏感词过滤，PHP实现的Trie树
- java采用 java.util.Properties类读取.properties配置文件
- DenseBox: Unifying Landmark Localization with End to End Object Detection
- 关于时间，日期，星期，月份的算法（Java中Calendar的使用方法）
- 递归练习
- Java 网络学习
- 敏感词检测
- Spring MVC 3.0.5+Spring 3.0.5+MyBatis3.0.4全注解实例详解（三）
- 中华武术
- 想学习做产品
- JAVA 按年月打印日历 Calendar的简单使用
