运行训练时报错:

IndexError: index 68 is out of bounds for axis 1 with size 68


解决办法:
不要删掉checkpoint文件夹里的yolov3_coco.ckpt的三个文件

妈蛋,好像不是这个原因,还是报错!!!

然后我尝试去除掉了训练文件中没有目标标注的空文件,还是不行:

去掉前:

去掉后:

后来查到些问题,貌似是没做数据清洗,标注的坐标是混乱的、有误的!!!

参考文章:IndexError: index 52 is out of bounds for axis 1 with size 52

作者在问题下提到:

后来我一查,果然坐标标注数据有问题:



所以,训练前必须做数据清洗!!!

还有,这么多同样的数据是咋回事,被病毒入侵了吗?


经过数据清洗,查到了一大堆坐标超限和错误class_id的错误数据:

D:\20200204_tensorflow_yolov3_train\python\python.exe D:/20200204_tensorflow_yolov3_train/tensorflow-yolov3/dontla_convert_annotation/Python_Convert_Annotations_From_Yolo_to_Tensorflow-Yolov3/convert_txt-Yolo2Tensorflow.py
错误标注:
12.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
389.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
754.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
862.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
1422.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
1479.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
1479.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
1831.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
2793.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
2803.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
2824.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3393.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3436.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3451.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3452.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3459.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3461.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3463.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3463.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3515.txt
[2, 0.446484, 0.140972, 0.146094, 0.151389]
[478, 47, 665, 156, 2]
错误标注:
3515.txt
[2, 0.492969, 0.500694, 0.139062, 0.206944]
[542, 286, 720, 435, 2]
错误标注:
3533.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3542.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3564.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3581.txt
[2, 0.344922, 0.509028, 0.122656, 0.170833]
[363, 305, 520, 428, 2]
错误标注:
3581.txt
[2, 0.734766, 0.234028, 0.146094, 0.179167]
[847, 104, 1034, 233, 2]
错误标注:
3581.txt
[2, 0.693359, 0.665278, 0.144531, 0.247222]
[795, 390, 980, 568, 2]
错误标注:
3581.txt
[2, 0.323047, 0.788889, 0.136719, 0.172222]
[326, 506, 501, 630, 2]
错误标注:
3604.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3610.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3615.txt
[2, 0.360547, 0.570139, 0.125781, 0.190278]
[381, 342, 542, 479, 2]
错误标注:
3615.txt
[2, 0.639844, 0.372917, 0.15625, 0.251389]
[719, 178, 919, 359, 2]
错误标注:
3673.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3758.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3779.txt
[2, 0.553516, 0.540278, 0.147656, 0.2]
[614, 317, 803, 461, 2]
错误标注:
3779.txt
[2, 0.349609, 0.757639, 0.138281, 0.206944]
[359, 471, 536, 620, 2]
错误标注:
3806.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3812.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3817.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3838.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3958.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
3986.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4040.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4055.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4087.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4091.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4109.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4138.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4144.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4164.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4188.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4223.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4284.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4318.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4348.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4357.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4362.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4376.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4397.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4397.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4421.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4426.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4443.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4460.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4466.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4472.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4486.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]
错误标注:
4662.txt
[0, 0.0, -0.138889, 0.0, 0.0]
[0, -100, 0, -100, 0]Process finished with exit code 0

将错误标注清除后,终于能正常训练了:



参考文章:将yolo标注转换为tensorflow_yolov3标注生成train.txt和test.txt同时做数据清洗

yunyang1994 tensorflow_yolov3训练报错:IndexError: index 68 is out of bounds for axis 1 with size 68 数据清洗相关推荐

  1. python报错:index 1 is out of bounds for axis 0 with size 1

    这里写自定义目录标题 解决方法 报错:index 1 is out of bounds for axis 0 with size 1 ** 解决方法 ** 不能像matlab一样省略一维数组前面的0( ...

  2. 成功解决IndexError: index 14 is out of bounds for axis 1 with size 14

    成功解决IndexError: index 14 is out of bounds for axis 1 with size 14 目录 解决问题 解决思路 解决方法 解决问题 IndexError: ...

  3. 成功解决IndexError: index 0 is out of bounds for axis 1 with size 0

    成功解决IndexError: index 0 is out of bounds for axis 1 with size 0 目录 解决问题 解决思路 解决方法 解决问题 IndexError: i ...

  4. IndexError: index 439 is out of bounds for axis 0 with size 439的解决办法

    IndexError: index 439 is out of bounds for axis 0 with size 439的解决办法: 在图像中遇到上述错误时,把xy的换一下即可.把第三行改为im ...

  5. 调用笔记本的摄像头实现基于opencv的视频人脸识别(中文显示和英文显示)以及 index 480 is out of bounds for axis 0 with size 480错误的解决

    @人脸识别代码和一些常见错误 基于opencv的视频人脸识别(中文显示)以及 index 480 is out of bounds for axis 0 with size 480错误的解决 参考了 ...

  6. mmdetection训练报错

    mmdetection训练报错. 先说结论吧,coco格式的annotations.json中categories的category_id不能有0(即背景类). 查了好几个小时,都快要疯了,才搞出这个 ...

  7. yolov5 v3.0训练报错: torch.nn.modules.module.ModuleAttributeError: ‘BatchNorm2d‘ object has no attribute

    欢迎大家关注笔者,你的关注是我持续更博的最大动力 原创文章,转载告知,盗版必究 yolov5 v3.0版本训练报错:torch.nn.modules.module.ModuleAttributeErr ...

  8. git报错error: index uses extension, fatal: index file corrupt

    git报错 error: index uses extension, which we do not understand fatal: index file corrupt 用git 拉取代码时,遇 ...

  9. Jenkins部署项目时,git报错fatal: index file smaller than expected

    @Jenkins部署项目时,git报错fatal: index file smaller than expectedTOC 最近在学习部署Jenkins时,微服务构建一直报错: Caused by: ...

最新文章

  1. 雷林鹏分享:jQuery EasyUI 数据网格 - 创建页脚摘要
  2. Step by Step WebMatrix网站开发之二:使用WebMatrix(2)
  3. Linux学习笔记——例说makefile 综合案例
  4. WebLogic启动与简单项目配置
  5. 基于jQuery日历插件制作日历
  6. exchange管理控制台命令行 查看邮箱数据库信息、接收连接器、发送连接器 相关命令
  7. HTMLTestRunner修改Python3的版本
  8. msf php脚本提权,[原创]WEB安全第六章 提权篇16 metasploit linux提权
  9. 实现网络IPv6平滑演进的DS-Lite CGN技术
  10. 把IE武装到牙齿,IE插件全攻略(转)
  11. 我常用的Windows软件
  12. 映美精相机(the imaging source)在树莓派(Raspberry Pi )中的使用(一)
  13. RAID-6技术详解
  14. win10系统电池图标不见了怎么恢复
  15. Elasticsearch 聚合搜索技术深入
  16. java中modifier_Ruby中的private modifier与Java中的对比
  17. 群同态和群同构的区别_同构和同态有什么区别,它们可以用在哪些方面?
  18. 电脑固态硬盘接口分类
  19. JS--JavaScript使用window对象操作框架集frameset中的各窗口(frames[])
  20. Win10微软补丁KB5018410致金蝶客户端无法连接中间件

热门文章

  1. java--callback
  2. webstorm安装_WebStorm2019安装教程
  3. 频谱仪使用方法图解_钳形电流表使用方法图解
  4. 【学习笔记】区块链应用判断准则
  5. 【实用】客户行项目清单FBL5N增加利润中心
  6. 【转载】扫盲概念RPA
  7. 【MM模块】 Goods Receipt 收货 1
  8. S/4HANA表现抢眼——第一季度SAP大中华区营收再现双位数增长
  9. 网上找到的一段关于SAP支持服务的QA (转)
  10. 利用BADI ME_PROCESS_PO_CUST進行PO check