[caffe]深度学习之CNN检测object detection方法摘要介绍

[caffe]深度学习之CNN检测object detection方法摘要介绍 

  3276人阅读  评论(1)  收藏  举报

一、RCNN

流程:

Extract region(off model) + extract features(on model) + classifyregions according feature (svm or softmax)

性能:

[caffe]深度学习之CNN检测object detection方法摘要介绍_第1张图片

精度:

[caffe]深度学习之CNN检测object detection方法摘要介绍_第2张图片

 

二、SPP-NET

流程:

先做conv,再根据window提取特征。为什么rcnn不能也这么做呢?原因在于spp对不同尺度进行了max pool处理能更好的满足不同尺度window的特征表达。

性能:

核心思想在全图只做一次conv,这个和overfeat的思想一致

 [caffe]深度学习之CNN检测object detection方法摘要介绍_第3张图片

精度:

[caffe]深度学习之CNN检测object detection方法摘要介绍_第4张图片

三、FAST-RCNN

流程:

引入了ROI层pooling,以及multi-task同时训练分类和检测框。

性能:

Compared to SPPnet, Fast R-CNN trains VGG163× faster, tests 10× faster, and is more accurate.

另外还额外提出了fc层SVD的思想

[caffe]深度学习之CNN检测object detection方法摘要介绍_第5张图片

Vgg时间性能分析

 [caffe]深度学习之CNN检测object detection方法摘要介绍_第6张图片

精度:

The improvement of Fast R-CNN over SPPnetillustrates that even though Fast R-CNN uses single-scale training and testing,fine-tuning the conv layers provides a large improvement in mAP (from 63.1% to66.9%). Traditional R-CNN achieves a mAP of 66.0%. These results arepragmatically valuable given how much faster and easier Fast R-CNN is to trainand test, which we discuss next.

 

 

 

 

四、FASTER-RCNN

流程:

在fast-rcnn的基础上,借鉴了FCN的思路,将proposal阶段转化成一个layer加进了网络一起学习。

[caffe]深度学习之CNN检测object detection方法摘要介绍_第7张图片

性能:

cost-free for proposal

精度:

our detection system has a frame rate of5fps (including all steps) on a GPU, while achieving state-of-the-art objectdetection accuracy on PASCAL VOC 2007 (73.2% mAP) and 2012 (70.4% mAP) using300 proposals per image


from: http://blog.csdn.net/sunbaigui/article/details/47728251

你可能感兴趣的:(DeepLearning,Caffe,ADAS_辅助驾驶_自动驾驶)