AlexeyAB / Yolo_mark

GUI for marking bounded boxes of objects in images for training neural network Yolo v3 and v2
https://github.com/AlexeyAB/darknet
The Unlicense
1.81k stars 682 forks source link

./yolo_mark.cmd #10

Open Abduoit opened 7 years ago

Abduoit commented 7 years ago

I have two issues First, when I run yolo_mark.cmd file I get the following error

abdulrahman@abdulrahman-ThinkPad-X230-Tablet:~/darknet/Yolo_mark/x64/Release$ ./yolo_mark.cmd Example how to start marking bouded boxes for training set Yolo v2 ./yolo_mark.cmd: line 4: yolo_mark.exe: command not found ./yolo_mark.cmd: line 7: pause: command not found

Second, When I want to train it, I use darknet.exe detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23 the file could be not found !!!

AlexeyAB commented 7 years ago
  1. Do you use Linux or Windows?

  2. Did you successfully compiled Yolo_mark by using MSVS2015 and OpenCV 2.4.x?

  3. Where is created file yolo_mark.exe, is there Yolo_mark/x64/Release/?

Abduoit commented 7 years ago

Thanks for reply @AlexeyAB 1- I am using Linux 2- Yes I compiled it correctly, using OpenCV 2.4.9. I saw the images of airplanes and birds

3- I don't have file yolo_mark.exe , there is file yolo_mark.cmd in /home/abdulrahman/Yolo_mark/x64/Release

AlexeyAB commented 7 years ago

From the manual: https://github.com/AlexeyAB/Yolo_mark#yolo_mark

To test, simply run

  • on Windows: x64/Release/yolo_mark.cmd
  • on Linux ./linux_mark.sh

So run ./linux_mark.sh

Abduoit commented 7 years ago

I already ran it, i saw my own images without bounding boxes, cuz I haven't trained yet,

I need to train it with my custom images, I think I did something wrong with the following step (3.2), I think I put the files in wrong directory. Could u tell me what does this mean near with executable darknet-file

3.2 Put files: yolo-obj.cfg, data/train.txt, data/obj.names, data/obj.data, darknet19_448.conv.23 and directory data/img near with executable darknet-file, and start training: darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23

So when I run this

darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23

I got

abdulrahman@abdulrahman-ThinkPad-X230-Tablet:~/Yolo_mark/x64/Release$ darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23 bash: darknet: command not found

AlexeyAB commented 7 years ago

I already ran it, i saw my own images without bounding boxes, cuz I haven't trained yet,

At first you should mark all of your images by ./linux_mark.sh


darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23

On linux you should run binary files in a such way: ./darknet instead of darknet

Abduoit commented 7 years ago

screenshot from 2017-07-13 16-52-03

should I run ./darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23 from ~/Yolo_mark/x64/Release$

As you see in the above pic there is no file ./darknet how can I run it

AlexeyAB commented 7 years ago

For Linux - you should download Darknet for Linux: https://github.com/pjreddie/darknet Compile it, find the runable file darknet, put near all files and run ./darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23

Abduoit commented 7 years ago

I have Darknet, it works properly So I put the runable file darknet as in this pic 11

I ran ./darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23 I got this pic 12

AlexeyAB commented 7 years ago

Put files: yolo-obj.cfg, data/train.txt, data/obj.names, data/obj.data, darknet19_448.conv.23 and directory data/img near with executable darknet-file, but not vice versa :)

https://github.com/AlexeyAB/Yolo_mark#yolo_mark

3.2 Put files: yolo-obj.cfg, data/train.txt, data/obj.names, data/obj.data, darknet19_448.conv.23 and directory data/img near with executable darknet-file, and start training: darknet detector train data/obj.data yolo-obj.cfg darknet19_448.conv.23

Also read this: https://github.com/AlexeyAB/darknet#how-to-train-to-detect-your-custom-objects

SheebanWasi commented 6 years ago

When I use to train coco-data sheebans-Air:data sheeban$ detector train cfg/coco.data cfg/yolov3.cfg darknet53.conv.74 -bash: detector: command not found I am in data directory because when I use sheebans-Air:data sheeban$ ./darknet detector train cfg/coco.data cfg/yolov3.cfg darknet53.conv.74 it throws error darknet directory not found

lalithkumawat commented 5 years ago
1. Do you use Linux or Windows?

2. Did you successfully compiled Yolo_mark by using MSVS2015 and OpenCV 2.4.x?

3. Where is created file `yolo_mark.exe`, is there `Yolo_mark/x64/Release/`?

not able to compile .sln in vs in windows :(

sky93 commented 4 years ago

Can anybody provide a .exe file?