Learning to Track at 100 FPS with Deep Regression Networks ECCV 2016 论文笔记
Learning to Track at 100 FPS with Deep Regression Networks ECCV 2016 论文笔记
工程网页:http://davheld.github.io/GOTURN/GOTURN.html
GitHub 地址:https://github.com/davheld/GOTURN
本文第一次利用 Deep Learning 技术将跟踪的速度维持在 100fps,当然是使用 GPU 的前提下。本文的流程框架如下所示:

将跟踪看做是回归问题,直接根据上一帧的位置,回归出当前帧的location。类比于 基于Siamese 网络的匹配,仅用第一帧作为 target object,本文方法不需要提候选的 proposal,直接进行 bounding box 的回归。很好的避开了 CNN 在跟踪问题上速度慢的难题。

是的,没了,就这么多。这就是文章的主要思想了。。。
另外:给些参考的blog,因为他们讲的更加详细。
1. http://blog.csdn.net/cuclxt/article/details/51570255
2. http://blog.csdn.net/autocyz/article/details/52648776
3. https://zhuanlan.zhihu.com/p/22715531 (强烈推荐)
Learning to Track at 100 FPS with Deep Regression Networks ECCV 2016 论文笔记的更多相关文章
- 《Hyperspectral Image Classification With Deep Feature Fusion Network》论文笔记
论文题目<Hyperspectral Image Classification With Deep Feature Fusion Network> 论文作者:Weiwei Song, Sh ...
- Deep Image Retrieval: Learning global representations for image search In ECCV, 2016学习笔记
- 论文地址:https://arxiv.org/abs/1604.01325 contribution is twofold: (i) we leverage a ranking framework ...
- 课程一(Neural Networks and Deep Learning),第四周(Deep Neural Networks)—— 0.学习目标
Understand the key computations underlying deep learning, use them to build and train deep neural ne ...
- Re-thinking Deep Residual Networks
本文是对ImageNet 2015的冠军ResNet(Deep Residual Networks)以及目前围绕ResNet这个工作研究者后续所发论文的总结,主要涉及到下面5篇论文. 1. Link: ...
- 课程一(Neural Networks and Deep Learning),第四周(Deep Neural Networks) —— 3.Programming Assignments: Deep Neural Network - Application
Deep Neural Network - Application Congratulations! Welcome to the fourth programming exercise of the ...
- Coursera Deep Learning 2 Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization - week1, Assignment(Initialization)
声明:所有内容来自coursera,作为个人学习笔记记录在这里. Initialization Welcome to the first assignment of "Improving D ...
- (转)Understanding, generalisation, and transfer learning in deep neural networks
Understanding, generalisation, and transfer learning in deep neural networks FEBRUARY 27, 2017 Thi ...
- Deep Learning论文笔记之(一)K-means特征学习
Deep Learning论文笔记之(一)K-means特征学习 zouxy09@qq.com http://blog.csdn.net/zouxy09 自己平时看了一些论文,但老感 ...
- This instability is a fundamental problem for gradient-based learning in deep neural networks. vanishing exploding gradient problem
The unstable gradient problem: The fundamental problem here isn't so much the vanishing gradient pro ...
随机推荐
- workerman 7272端口被占用
1/问题:workerman 7272端口被占用 2/策略: 1.查找被占用的端口 netstat -tln netstat -tln | grep 8083 netstat -tln 查看端口使用情 ...
- 08 集合[11,22,33,44,55,66,77,88,99],将所有<66的值保存至字典的第一个key中,将所有>=66的值保存至字典的第二个key中。即:{'k1':<66的所有值,'k2':>=66的所有值}
li = [11,22,33,44,55,66,77,88,99]dict = {'k1':[],'k2':[]}for i in li: if i < 66: dict[& ...
- [博客迁移]探索Windows Azure 监控和自动伸缩系列3 - 启用Azure监控扩展收集自定义监控数据
上一篇我们介绍了获取Azure的监控指标和监控数据: http://www.cnblogs.com/teld/p/5113376.html 本篇我们继续:监控虚拟机的自定义性能计数器. 随着我们应用规 ...
- caffe生成voc格式lmdb
要训练ssd基本都是在liu wei框架下改,生成lmdb这一关照葫芦画瓢总遇坑,记录之: 1. labelmap_voc.prototxt要根据自己的分类修改,比如人脸检测改成这样: item { ...
- ArrayList 底层实现原理
ArrayList的底层实现原理 1, 属性:private static final int DEFAULT_CAPACITY = 10; private static final Object [ ...
- POJ 3662 Telephone Lines (二分 + 最短路)
Farmer John wants to set up a telephone line at his farm. Unfortunately, the phone company is uncoop ...
- Django后端项目---- rest framework(4)
一.分页 试问如果当数据量特别大的时候,你是怎么解决分页的? 方式a.记录当前访问页数的数据id 方式b.最多显示120页等 方式c.只显示上一页,下一页,不让选择页码,对页码进行加密 1.基于lim ...
- Java中使用自定义类封装数组,添加类方法实现数据操作
1.具体见注释 2.后续或有更新 public class MyArray { private long[] array; private int cnt; // 自定义数组类的元素个数 /** 使用 ...
- 最简单的uwsgi+nginx配置多个django站点
1. nginx.conf http{ server { listen 80; server_name www.web1.com ....... location / { uwsgi_p ...
- Golang 数组和字符串之间的相互转换[]byte/string
package main import ( "fmt" ) func main() { str := "hello" arr := []byte(str) fm ...