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

  1. 《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 ...

  2. 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 ...

  3. 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 ...

  4. 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: ...

  5. 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 ...

  6. 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 ...

  7. (转)Understanding, generalisation, and transfer learning in deep neural networks

    Understanding, generalisation, and transfer learning in deep neural networks FEBRUARY 27, 2017   Thi ...

  8. 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 ...

  9. 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

  1. 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 ...

  2. 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[& ...

  3. [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 ...

  4. 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 { ...

  5. 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 [ ...

  6. 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 ...

  7. 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 ...

  8. 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 ...

  9. 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 ...

  10. 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 ...