classification    ./examples/cifar10/cifar10_full.prototxt  ./examples/cifar10/cifar10_full_iter_70000.caffemodel.h5 ./examples/cifar10/mean.binaryproto    ./examples/cifar10/labels.txt   ~/Downloads/images/horse/.jpg

sea@sea-X550JK:~/caffe$ classification --help
Usage: classification deploy.prototxt network.caffemodel mean.binaryproto labels.txt img.jpg classification models/bvlc_reference_caffenet/deploy.prototxt
models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel
models/bvlc_reference_caffenet/mean.binaryproto
models/bvlc_reference_caffenet/labels.txt
~/Downloads/images/horse/.jpg

用cifar10训练的结果进行分类:  

python python/classify.py --model_def examples/cifar10/cifar10_quick.prototxt --pretrained_model examples/cifar10/cifar10_quick_iter_5000.caffemodel.h5 --center_only  examples/images/cat.jpg foo
python python/classify.py --model_def models/bvlc_reference_caffenet/deploy.prototxt --pretrained_model models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel --center_only  examples/images/cat.jpg foo
I1103 16:59:58.189568 25346 net.cpp:200] conv1 does not need backward computation.
I1103 16:59:58.189571 25346 net.cpp:200] data does not need backward computation.
I1103 16:59:58.189574 25346 net.cpp:242] This network produces output prob
I1103 16:59:58.189584 25346 net.cpp:255] Network initialization done.
I1103 16:59:58.303480 25346 upgrade_proto.cpp:44] Attempting to upgrade input file specified using deprecated transformation parameters: models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel
I1103 16:59:58.303509 25346 upgrade_proto.cpp:47] Successfully upgraded file specified using deprecated data transformation parameters.
W1103 16:59:58.303514 25346 upgrade_proto.cpp:49] Note that future Caffe releases will only support transform_param messages for transformation fields.
I1103 16:59:58.303517 25346 upgrade_proto.cpp:53] Attempting to upgrade input file specified using deprecated V1LayerParameter: models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel
I1103 16:59:58.504439 25346 upgrade_proto.cpp:61] Successfully upgraded file specified using deprecated V1LayerParameter
I1103 16:59:58.559579 25346 net.cpp:744] Ignoring source layer loss
/usr/local/lib/python2.7/dist-packages/skimage/transform/_warps.py:84: UserWarning: The default mode, 'constant', will be changed to 'reflect' in skimage 0.15.
warn("The default mode, 'constant', will be changed to 'reflect' in "
Loading file: examples/images/cat.jpg
Classifying 1 inputs.
Done in 1.20 s.
Predictions : [[ 7.96905475e-09 2.68402800e-05 4.61699550e-08 5.81401345e-08
3.00355154e-08 1.08543240e-07 7.21305184e-08 6.65618529e-07
4.10124194e-05 8.26508540e-07 2.64434061e-06 4.29981719e-06
2.29038033e-05 9.16178294e-07 2.02221463e-06 1.91530648e-06
8.36403979e-06 5.25011237e-05 1.32120860e-07 7.34086640e-08
7.26202700e-07 6.55063502e-07 2.83661024e-07 8.35531750e-08
1.45248293e-07 3.21299929e-08 5.94506417e-08 1.11880944e-07
2.61020752e-08 1.33058847e-05 2.00340565e-07 7.72992621e-08
2.47393245e-07 5.60683127e-08 7.26820346e-08 2.93914972e-08
8.09441403e-08 1.17543671e-07 1.24727379e-07 1.14408145e-07
sea@sea-X550JK:~/caffe$ python  readFromFooAndShow.py
sz = 4112
nl.shape = (1, 1000)
ssdict = [(281, 0.30427486), (285, 0.1783575), (282, 0.16652611), (287, 0.15713461), (278, 0.042343788), (277, 0.039970074),
(283, 0.011617188), (876, 0.0085467361), (284, 0.0076080137), (463, 0.0066294265), (904, 0.0065242196), (968, 0.0063064895),
(259, 0.0051229554), (330, 0.0046631121), (760, 0.0044421358), (478, 0.0042510382), (331, 0.0039331503), (728, 0.003812969),
(280, 0.0035846629), (588, 0.0033092475), (861, 0.0028945252), (332, 0.0026644215), (333, 0.0022166823), (151, 0.0021597522),
(356, 0.0018406865), (552, 0.0016959301), (435, 0.00094394217), (896, 0.00084631733), (937, 0.00082845741), (335, 0.00076790486),
(897, 0.0007364807), (519, 0.00072649814), (674, 0.00063642312), (457, 0.00062823156), (263, 0.00055513595), (969, 0.00043508445),
(773, 0.00041424474), (794, 0.00039454823), (230, 0.00037321725), (534, 0.00036081325), (104, 0.00032497221), (272, 0.00032023937),
(473, 0.0003057541), (725, 0.00030245754), (742, 0.00029926837), (722, 0.00028606801), (987, 0.00024712173), (622, 0.00024177019),
(274, 0.00023734267),

下面是分类的过程bvlc_reference_caffenet:

模型bvlc_reference_caffenet 是用于分类的:

./build/examples/cpp_classification/classification.bin \
models/bvlc_reference_caffenet/deploy.prototxt \
models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel \
data/ilsvrc12/imagenet_mean.binaryproto \
data/ilsvrc12/synset_words.txt \
examples/images/cat.jpg
sea@sea-X550JK:~/caffe$ ./build/examples/cpp_classification/classification.bin \
> models/bvlc_reference_caffenet/deploy.prototxt \
> models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel \
> data/ilsvrc12/imagenet_mean.binaryproto \
> data/ilsvrc12/synset_words.txt \
> examples/images/cat.jpg
---------- Prediction for examples/images/cat.jpg ----------
0.3134 - "n02123045 tabby, tabby cat"
0.2380 - "n02123159 tiger cat"
0.1235 - "n02124075 Egyptian cat"
0.1003 - "n02119022 red fox, Vulpes vulpes"
0.0715 - "n02127052 lynx, catamount"

预测的实例/图像/————————cat.jpg
“n02123045 46 6猫,虎斑猫”
“n02123159 0.2380老虎猫”
“n02124075 0.1235埃及猫”
“n02119022 0.1003赤狐,狐狐”
“n02127052猞猁,0.0715美洲豹”

./build/examples/cpp_classification/classification.bin \
models/bvlc_reference_caffenet/deploy.prototxt \
models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel \
data/ilsvrc12/imagenet_mean.binaryproto \
data/ilsvrc12/synset_words.txt \
/home/sea/Downloads/images/person.jpeg
/home/sea/Downloads/images/person.jpeg 

---------- Prediction for /home/sea/Downloads/images/person.jpeg ----------
0.8322 - "n04350905 suit, suit of clothes"
0.0799 - "n04591157 Windsor tie"
0.0588 - "n02883205 bow tie, bow-tie, bowtie"
0.0051 - "n10148035 groom, bridegroom"
0.0041 - "n02865351 bolo tie, bolo, bola tie, bola"

“n04350905 0.8322服,服之衣”
“n04591157 0.0799领带。”
“n02883205 0.0588蝴蝶结领带,领结,bowtie”
“n10148035马夫,bridegroom率”
“n02865351联络0.0041蛋糕,蛋糕,球铁,球”

识别装修图片:

./build/examples/cpp_classification/classification.bin \
models/bvlc_reference_caffenet/deploy.prototxt \
models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel \
data/ilsvrc12/imagenet_mean.binaryproto \
data/ilsvrc12/synset_words.txt \
/home/sea/Downloads/images/a.jpg

>   /home/sea/Downloads/images/a.jpg
---------- Prediction for /home/sea/Downloads/images/a.jpg ----------
0.3274 - "n04081281 restaurant, eating house, eating place, eatery"
0.1335 - "n03761084 microwave, microwave oven"
0.1196 - "n03661043 library"
0.0768 - "n04553703 washbasin, handbasin, washbowl, lavabo, wash-hand basin"
0.0710 - "n03742115 medicine chest, medicine cabinet"
0.3274“n04081281餐厅,吃房子,吃的地方,餐馆”
0.1335“n03761084微波,微波炉”
0.1196“n03661043图书馆”
0.0768“n04553703洗脸盆,洗手盆,洗脸盆,洗手盆,洗手盆”
0.0710“n03742115药箱,药箱”

目标检测、定位的+目标识别的fetch_faster_rcnn_models:

https://github.com/rbgirshick/py-faster-rcnn/blob/master/data/scripts/fetch_faster_rcnn_models.sh

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

Download pre-computed Faster R-CNN detectors cd $FRCN_ROOT
./data/scripts/fetch_faster_rcnn_models.sh This will populate the $FRCN_ROOT/data folder with faster_rcnn_models. See data/README.md for details. These models were trained on VOC 2007 trainval.

ref https://github.com/rbgirshick/py-faster-rcnn/blob/master/data/scripts/fetch_faster_rcnn_models.sh

目标检测--resnet-50:

./build/examples/cpp_classification/classification.bin \
/media/sea/wsWin10/wsWindows10/ws_caffe/model-zoo/ResNet-50/deploy.prototxt \
/media/sea/wsWin10/wsWindows10/ws_caffe/model-zoo/ResNet-50/ResNet-50-model.caffemodel \
data/ilsvrc12/imagenet_mean.binaryproto \
data/ilsvrc12/synset_words.txt \
/home/sea/Downloads/images/a.jpg

人脸识别的:

caffe学习--cifar10学习-ubuntu16.04-gtx650tiboost--1g--03--20171103的更多相关文章

  1. caffe学习--cifar10学习-ubuntu16.04-gtx650tiboost--1g--02

    caffe学习--cifar10学习-ubuntu16.04-gtx650tiboost--1g--02 训练网络: caffe train -solver examples/cifar10/cifa ...

  2. caffe学习一:ubuntu16.04下跑Faster R-CNN demo (基于caffe). (亲测有效,记录经历两天的吐血经历)

    兜兜转转,兜兜转转; 一次有一次,这次终于把Faster R-CNN 跑通了. 重要提示1:在开始跑Faster R-CNN之前一定要搞清楚用的是Python2 还是Python3. 不然你会无限次陷 ...

  3. 深度学习环境配置Ubuntu16.04+CUDA8.0+CUDNN5

    深度学习从12年开始打响,配置深度学习环境软件一直是一个头疼的问题,如何安装显卡驱动,如何安装CUDA,如何安装CUDNN:Ubuntu官方一直吐槽Nvidia显卡驱动有问题,网上大神也给出了关闭li ...

  4. 深度学习环境配置:Ubuntu16.04安装GTX1080Ti+CUDA9.0+cuDNN7.0完整安装教程(多链接多参考文章)

    本来就对Linux不熟悉,经过几天惨痛的教训,参考了不知道多少篇文章,终于把环境装好了,每篇文章或多或少都有一些用,但没有一篇完整的能解决我安装过程碰到的问题,所以决定还是自己写一篇我安装过程的教程, ...

  5. 深度学习环境配置:Ubuntu16.04下安装GTX1080Ti+CUDA9.0+cuDNN7.0完整安装教程(多链接多参考文章)

    本来就对Linux不熟悉,经过几天惨痛的教训,参考了不知道多少篇文章,终于把环境装好了,每篇文章或多或少都有一些用,但没有一篇完整的能解决我安装过程碰到的问题,所以决定还是自己写一篇我安装过程的教程, ...

  6. caffe学习--cifar10学习-ubuntu16.04-gtx650tiboost--1g--01

    引用了下文的资料,在此感谢! http://www.cnblogs.com/alexcai/p/5468164.html http://blog.csdn.net/garfielder007/arti ...

  7. ROS入门学习(基于Ubuntu16.04+kinetic)

    本文主要部分全部来源于ROS官网的Tutorials. Setup roscore # making sure that we have roscore running rosrun turtlesi ...

  8. Ubuntu16.04+cuda8.0rc+opencv3.1.0+caffe+Theano+torch7搭建教程

    https://blog.csdn.net/jywowaa/article/details/52263711 学习中用到深度学习的框架,需要搭建caffe.theano和torch框架.经过一个月的不 ...

  9. win10安装ubuntu16.04及后续配置

    原文地址:https://www.jianshu.com/p/842e36a8255c UEFI 模式下win10安装ubuntu16.04双系统教程 - baobei0112的专栏 - CSDN博客 ...

随机推荐

  1. CSS3 基本属性 浅析(含选择器、背景阴影、3D转换、动画等)

    1渐进增强原则 2私有前缀  不同浏览器在发布不同版本(一般测试版)时会加前缀,新增属性加上前缀进行支持测试:     Chrome浏览器:-webkit-border-radius: 5px;   ...

  2. 虚拟 ​router 原理分析

    上一节我们创建了虚拟路由器“router_100_101”,并通过 ping 验证了 vlan100 和 vlan101 已经连通. 本节将重点分析其中的原理. 首先我们查看控制节点的 linux b ...

  3. 洛谷P1103 书本整理

    题目描述 Frank是一个非常喜爱整洁的人.他有一大堆书和一个书架,想要把书放在书架上.书架可以放下所有的 书,所以Frank首先将书按高度顺序排列在书架上.但是Frank发现,由于很多书的宽度不同, ...

  4. JDBC连接池&DBUtils使用

    使用连接池改造JDBC的工具类: 1.1.1          需求: 传统JDBC的操作,对连接的对象销毁不是特别好.每次创建和销毁连接都是需要花费时间.可以使用连接池优化的程序. * 在程序开始的 ...

  5. 转 C/C++内存分配方式与存储区

    C/C++内存分配方式与存储区 C/C++内存分配有三种方式:[1]从静态存储区域分配.内存在程序编译的时候就已经分配好,这块内存在程序的整个运行期间都存在.例如全局变量,static变量.[2]在栈 ...

  6. 《Linux命令行与shell脚本编程大全 第3版》Linux命令行---23

    以下为阅读<Linux命令行与shell脚本编程大全 第3版>的读书笔记,为了方便记录,特地与书的内容保持同步,特意做成一节一次随笔,特记录如下:

  7. LeetCode OJ--Sum Root to Leaf Numbers

    https://oj.leetcode.com/problems/sum-root-to-leaf-numbers/ 给一棵树,找从根到叶的路径,并把路径上的数相加,求和. 树的深度优先搜索 /** ...

  8. Android判断是否为刘海屏

    主要总结主流品牌小米.华为.oppo.vivo的刘海屏判断.在某些特殊页面需要适配刘海屏时,可以用以下方法判断.或者判断屏幕比例是否大于2. /** * 小米刘海屏判断. */ public stat ...

  9. function in Postgres

    CREATE or REPLACE FUNCTION fn_attr_category() RETURNS void AS $BODY$ declare v_tmp_rec record; begin ...

  10. Codeforces 766E Mahmoud and a xor trip(树形DP)

    题目链接 Mahmoud and a xor trip 树形DP.先考虑每个点到他本身的距离和,再算所有点两两距离和. 做的时候考虑二进制拆位即可. #include <bits/stdc++. ...