目标检测--Focal Loss for Dense Object Detection
来源:互联网 发布:镧系元素物理性质知乎 编辑:程序博客网 时间:2024/05/22 17:40
Focal Loss for Dense Object Detection
ICCV2017
https://arxiv.org/abs/1708.02002
本文算是用简单的方法解决复杂的问题了,好的思想是简单的
针对目标检测,目前有两类主流算法: two-stage detectors 和 one-stage detectors, two-stage detectors 的精度好但是速度慢,one-stage detectors 速度快精度差一些,本文指出one-stage detectors 精度差主要是因为 在训练时的 class imbalance,不仅仅是正负样本不平衡,更主要的是难易样本比例严重失调。随后提出了改进损失函数的定义,减小大量简单背景样本对总体损失函数的贡献,相对提高难得样本在损失函数中的权重。
In this work, we identify class imbalance as the primary obstacle preventing one-stage object detectors from surpassing top-performing, two-stage methods, such as Faster R-CNN variants. To address this, we propose the focal loss which applies a modulating term to the cross entropy loss in
order to focus learning on hard examples and down-weight the numerous easy negatives.
- Related Work
先看看目标检测 的 历史
Classic Object Detectors: 最经典的思路就会 滑动窗口了,Adboost 用于人脸检测, HOG, DPMs 等
Two-stage Detectors: 首先是 候选区域提取,然后是使用 CNN 进行分类,从最开始的R-CNN,Fast R-CNN, 到最经典的框架就是 Faster R-CNN framework
One-stage Detectors: One stage detectors are applied over a regular, dense sampling of object locations, scales, and aspect ratios
代表性方法:OverFeat, SSD , YOLO
Class Imbalance: 不管是传统的 one-stage object detection 方法 如 boosted detectors , DPMs 还是 最近的 SSD,在训练阶段都面临一个很大的 class imbalance, 这些检测器在一幅图像中大约要 评估 10000-100000 个候选位置,但是只有很少的位置含有物体。 这个不平衡会导致两个问题:1)训练的低效率,因为大部分位置都是简单的负样本,他们没有什么有用的学习信息。2)简单负样本占整个样本的绝大多数,导致学习到的模型泛化性能降低。 以前解决这个问题的方法一般是 hard negative mining 或者赋予不同权重的策略 more complex sampling/reweighing schemes。
本文提出的 focal loss 很好的解决了 class imbalance,可以高效的训练所有的样本,不用设计采样策略来减少简单的负样本
- Focal Loss
我们首先从 二分类问题中的 cross entropy (CE) loss 谈起,慢慢引入 focal loss。
换一个马夹是这样的:
One notable property of this loss, which can be easily seen in its plot, is that even examples that are easily classified (p t>.5) incur a loss with non-trivial magnitude. When summed over a large number of easy examples, these small loss values can overwhelm the rare class
大量简单的负样本 对总体的损失函数影响太大
3.1. Balanced Cross Entropy
解决这个class imbalance 一个常规方法就是引入 a weighting factor α
3.2. Focal Loss Definition
focal loss 有两个属性:
1) 当一个样本被误分类, p_t 很小,误分类引入的误差不受影响 , the modulating factor is near 1 and the loss is unaffected
2) The focusing parameter γ smoothly adjusts the rate at which easy examples are down-weighted 。
The focusing parameter γ 会平滑的调整 降低简单负样本的权重
实际使用的 focal loss 引入了 α-balanced
3.3. Class Imbalance and Model Initialization
class imbalance 会导致 训练初期的不稳定,这里我们引入了 先验知识, 一般为 0.01
the value of p estimated byt he model for the rare class (foreground) at the start of training
3.4. Class Imbalance and Two-stage Detectors
Two-stage Detectors 是怎么解决 class imbalance 了?
(1) a two-stage cascade and (2) biased minibatch sampling, 1:3
- RetinaNet Detector
这里我们设计了一个 RetinaNet Detector 来验证我们提出的 local loss 的有效性
Feature Pyramid Network (FPN) + subnetworks for classifying anchor boxes + subnetworks for anchor boxes regress
- 目标检测--Focal Loss for Dense Object Detection
- 目标检测“Focal Loss for Dense Object Detection”
- Focal Loss for Dense Object Detection
- Focal Loss for Dense Object Detection
- Focal Loss for Dense Object Detection
- Focal Loss for Dense Object Detection
- Focal Loss for Dense Object Detection
- 论文阅读-《Focal Loss for Dense Object Detection》
- 论文解读Focal Loss for Dense Object Detection
- [论文阅读] Focal Loss for Dense Object Detection
- [论文笔记]Focal Loss for Dense Object Detection
- 《Focal Loss for Object Detection》阅读笔记
- 目标检测之focal loss
- 目标检测“A MultiPath Network for Object Detection”
- 目标检测--Feature Pyramid Networks for Object Detection
- 目标检测“Feature Pyramid Networks for Object Detection”
- 目标检测“Perceptual Generative Adversarial Networks for Small Object Detection”
- 目标检测“Perceptual Generative Adversarial Networks for Small Object Detection”
- ovirt-node 之build分析
- 2018 网易
- Centos 7.0下面安装Mysql5.7.18
- 习题 6.5 将一个数组中的值按逆序重新存放。例如,原来顺序为8,6,5,4,1。要求改为1,4,5,6,8。
- Hydra 8.4/8.5新增功能
- 目标检测--Focal Loss for Dense Object Detection
- python 实现全角半角的相互转换
- 通过bat脚本启动关闭oracle实例
- chown 文件拥有者
- python中time模块与datatime模块用法
- Java 8的新特性—终极版
- yii 百度编辑器Ueditor使用
- nodejs搭建静态服务器 用Nodejs搭建服务器访问html、css、js等外部链接
- 给Java新手的一些建议----Java知识点归纳(Java基础部分)