darknet yolov3训练自己的数据集出现的问题

最新推荐文章于 2024-06-17 15:52:48 发布

980205

最新推荐文章于 2024-06-17 15:52:48 发布

阅读量1.6k

点赞数

分类专栏：目标检测

本文链接：https://blog.csdn.net/hyl999/article/details/90765633

版权

目标检测专栏收录该内容

11 篇文章 0 订阅

订阅专栏

之前步骤都是在window准备好的，给图片打标签，将xml转txt等，现在在linux训练

执行训练命令：

./darknet detector train cfg/yolov3-mouth.data cfg/yolov3-tiny-mouth.cfg

出现如下错误：不能加载图片和txt文件

wl@ubuntu:/mnt/hgfs/kshare/darknet/darknet-master$ ./darknet detector train cfg/yolov3-mouth.data cfg/yolov3-tiny-mouth.cfg
yolov3-tiny-mouth
batch: Using default '1'
subdivisions: Using default '1'
layer     filters    size              input                output
    0 conv     16  3 x 3 / 1   416 x 416 x   3   ->   416 x 416 x  16  0.150 BFLOPs
    1 max          2 x 2 / 2   416 x 416 x  16   ->   208 x 208 x  16
    2 conv     32  3 x 3 / 1   208 x 208 x  16   ->   208 x 208 x  32  0.399 BFLOPs
    3 max          2 x 2 / 2   208 x 208 x  32   ->   104 x 104 x  32
    4 conv     64  3 x 3 / 1   104 x 104 x  32   ->   104 x 104 x  64  0.399 BFLOPs
    5 max          2 x 2 / 2   104 x 104 x  64   ->    52 x  52 x  64
    6 conv    128  3 x 3 / 1    52 x  52 x  64   ->    52 x  52 x 128  0.399 BFLOPs
    7 max          2 x 2 / 2    52 x  52 x 128   ->    26 x  26 x 128
    8 conv    256  3 x 3 / 1    26 x  26 x 128   ->    26 x  26 x 256  0.399 BFLOPs
    9 max          2 x 2 / 2    26 x  26 x 256   ->    13 x  13 x 256
   10 conv    512  3 x 3 / 1    13 x  13 x 256   ->    13 x  13 x 512  0.399 BFLOPs
   11 max          2 x 2 / 1    13 x  13 x 512   ->    13 x  13 x 512
   12 conv   1024  3 x 3 / 1    13 x  13 x 512   ->    13 x  13 x1024  1.595 BFLOPs
   13 conv    256  1 x 1 / 1    13 x  13 x1024   ->    13 x  13 x 256  0.089 BFLOPs
   14 conv    512  3 x 3 / 1    13 x  13 x 256   ->    13 x  13 x 512  0.399 BFLOPs
   15 conv     18  1 x 1 / 1    13 x  13 x 512   ->    13 x  13 x  18  0.003 BFLOPs
   16 yolo
   17 route  13
   18 conv    128  1 x 1 / 1    13 x  13 x 256   ->    13 x  13 x 128  0.011 BFLOPs
   19 upsample            2x    13 x  13 x 128   ->    26 x  26 x 128
   20 route  19 8
   21 conv    256  3 x 3 / 1    26 x  26 x 384   ->    26 x  26 x 256  1.196 BFLOPs
   22 conv     18  1 x 1 / 1    26 x  26 x 256   ->    26 x  26 x  18  0.006 BFLOPs
   23 yolo
Learning Rate: 0.001, Momentum: 0.9, Decay: 0.0005
Resizing
512
"annot load image "/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-244.jpg
STB Reason: can't fopen

我猜测可能是路径不对，但我用opencv又可以读取路径所在的文件，那有些奇怪了，又来有把图片放到darknet下的data目录下，也一样的出现这问题，搞了两个小时，无语了。

后来github上查找，原来原因是我在window下生成的txt文件，是如下图解释：

原来是在window生成txt文件，换行都是 \r\n 来结束，而linux是\n结束，而我是在window生成的txt，在linux下训练，导致图片路径加载不对，那需要解决把\r\n替换成\n

解决方法：在linux 下用命令使用"cat -A [Filename]" 查看，如下图所示，看到的为一个Windows形式的换行符，\r对应符号^M，\n对应符号$. 结果如下：

wl@ubuntu:/mnt/hgfs/kshare/darknet/darknet-master$ cat -A data/mouth_train.txt 
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-170.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-171.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-172.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-173.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-174.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-175.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-176.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-177.jpg^M$
/mnt/hgfs/kshare/darknet/darknet-master/data/image/SCUT-FBP-178.jpg^M$

再使用命令将 \r\n转换成\n, 使用命令如下：

wl@ubuntu:/mnt/hgfs/kshare/darknet/darknet-master$ sed -i 's/\r//g' data/mouth_test.txt 
wl@ubuntu:/mnt/hgfs/kshare/darknet/darknet-master$ sed -i 's/\r//g' data/mouth_test.txt

再执行训练命令，终于可以了：

wl@ubuntu:/mnt/hgfs/kshare/darknet/darknet-master$ ./darknet detector train cfg/yolov3-mouth.data cfg/yolov3-tiny-mouth.cfg
yolov3-tiny-mouth
batch: Using default '1'
subdivisions: Using default '1'
layer     filters    size              input                output
    0 conv     16  3 x 3 / 1   416 x 416 x   3   ->   416 x 416 x  16  0.150 BFLOPs
    1 max          2 x 2 / 2   416 x 416 x  16   ->   208 x 208 x  16
    2 conv     32  3 x 3 / 1   208 x 208 x  16   ->   208 x 208 x  32  0.399 BFLOPs
    3 max          2 x 2 / 2   208 x 208 x  32   ->   104 x 104 x  32
    4 conv     64  3 x 3 / 1   104 x 104 x  32   ->   104 x 104 x  64  0.399 BFLOPs
    5 max          2 x 2 / 2   104 x 104 x  64   ->    52 x  52 x  64
    6 conv    128  3 x 3 / 1    52 x  52 x  64   ->    52 x  52 x 128  0.399 BFLOPs
    7 max          2 x 2 / 2    52 x  52 x 128   ->    26 x  26 x 128
    8 conv    256  3 x 3 / 1    26 x  26 x 128   ->    26 x  26 x 256  0.399 BFLOPs
    9 max          2 x 2 / 2    26 x  26 x 256   ->    13 x  13 x 256
   10 conv    512  3 x 3 / 1    13 x  13 x 256   ->    13 x  13 x 512  0.399 BFLOPs
   11 max          2 x 2 / 1    13 x  13 x 512   ->    13 x  13 x 512
   12 conv   1024  3 x 3 / 1    13 x  13 x 512   ->    13 x  13 x1024  1.595 BFLOPs
   13 conv    256  1 x 1 / 1    13 x  13 x1024   ->    13 x  13 x 256  0.089 BFLOPs
   14 conv    512  3 x 3 / 1    13 x  13 x 256   ->    13 x  13 x 512  0.399 BFLOPs
   15 conv     18  1 x 1 / 1    13 x  13 x 512   ->    13 x  13 x  18  0.003 BFLOPs
   16 yolo
   17 route  13
   18 conv    128  1 x 1 / 1    13 x  13 x 256   ->    13 x  13 x 128  0.011 BFLOPs
   19 upsample            2x    13 x  13 x 128   ->    26 x  26 x 128
   20 route  19 8
   21 conv    256  3 x 3 / 1    26 x  26 x 384   ->    26 x  26 x 256  1.196 BFLOPs
   22 conv     18  1 x 1 / 1    26 x  26 x 256   ->    26 x  26 x  18  0.006 BFLOPs
   23 yolo
Learning Rate: 0.001, Momentum: 0.9, Decay: 0.0005
Resizing
352
Loaded: 0.099826 seconds
Region 16 Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.508808, .5R: -nan, .75R: -nan,  count: 0
Region 23 Avg IOU: 0.172283, Class: 0.671358, Obj: 0.755280, No Obj: 0.487993, .5R: 0.000000, .75R: 0.000000,  count: 1
1: 254.888275, 254.888275 avg, 0.000000 rate, 9.877372 seconds, 1 images
Loaded: 0.000069 seconds
Region 16 Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.508524, .5R: -nan, .75R: -nan,  count: 0
Region 23 Avg IOU: 0.098354, Class: 0.338703, Obj: 0.193994, No Obj: 0.488208, .5R: 0.000000, .75R: 0.000000,  count: 1
2: 260.959106, 255.495361 avg, 0.000000 rate, 2.503628 seconds, 2 images
Loaded: 0.000075 seconds
Region 16 Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.510984, .5R: -nan, .75R: -nan,  count: 0
Region 23 Avg IOU: 0.284477, Class: 0.340647, Obj: 0.187216, No Obj: 0.490305, .5R: 0.000000, .75R: 0.000000,  count: 1
3: 256.483185, 255.594147 avg, 0.000000 rate, 2.478287 seconds, 3 images
Loaded: 0.018822 seconds
Region 16 Avg IOU: 0.163654, Class: 0.443973, Obj: 0.593278, No Obj: 0.511138, .5R: 0.000000, .75R: 0.000000,  count: 1
Region 23 Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.492911, .5R: -nan, .75R: -nan,  count: 0
4: 264.743774, 256.509094 avg, 0.000000 rate, 2.654358 seconds, 4 images
Loaded: 0.000085 seconds
Region 16 Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.507896, .5R: -nan, .75R: -nan,  count: 0
Region 23 Avg IOU: 0.174168, Class: 0.560516, Obj: 0.354948, No Obj: 0.493261, .5R: 0.000000, .75R: 0.000000,  count: 1
5: 262.642822, 257.122467 avg, 0.000000 rate, 2.555897 seconds, 5 images
Loaded: 0.000080 seconds
Region 16 Avg IOU: 0.070988, Class: 0.595818, Obj: 0.754916, No Obj: 0.509035, .5R: 0.000000, .75R: 0.000000,  count: 1
Region 23 Avg IOU: -nan, Class: -nan, Obj: -nan, No Obj: 0.491065, .5R: -nan, .75R: -nan,  count: 0
6: 267.453857, 258.155609 avg, 0.000000 rate, 2.463075 seconds, 6 images
Loaded: 0.000143 seconds

980205

关注

0
点赞
踩
9

收藏

觉得还不错? 一键收藏
3
评论
darknet yolov3训练自己的数据集出现的问题

之前步骤都是在window准备好的，给图片打标签，将xml转txt等，现在在linux训练执行训练命令：./darknet detector train cfg/yolov3-mouth.data cfg/yolov3-tiny-mouth.cfg出现如下错误：不能加载图片和txt文件wl@ubuntu:/mnt/hgfs/kshare/darknet/darknet-maste...
复制链接

扫一扫

专栏目录