file directory

[Darienzhangyy]

Hi,I am new to object detection and have been struggling on training Yolo on my data for the past few days. So I would really appreciate any suggestion.

I am able to generate the files like "after_conversion.txt" and "training_list.txt". But I am confused about the path I should create.

Here are my files: ls ~/caffe/darknet/scripts voc_label.py WISCR WISCR_label.py

Here WISCR_label.py is like your convert.py. I am able to run this script but not sure where I should save the output file.

ls WISCR/data annotations Annotations images Images ImageSets labels train.mat

I created my data following the format of PASCAL VOC files, with "Annotations"(a bunch of xml files), "Images"(original JPG) and "ImageSets"(train.txt storing the absolute paths of the images).

"labels" stores the output( ) from the WISCR_label.py.(called train.txt)

I know it is quite important to modify the src/yolo.c file and the path should point to the output file of WISCR_label.py. But what should be in the same parent directory as the "train.txt"? And should I make a "annotations" folder and a "images"(lower-cases) folder in my WISCR/data folder?

Thank you so much.

[Darienzhangyy]

And here is the error message. May be this is because my input data directory is not right... root@ip-172-31-10-52:~/caffe/darknet# ./darknet yolo train cfg/yolo.cfg extraction.conv.weights yolo 0: Convolutional Layer: 448 x 448 x 3 image, 64 filters -> 224 x 224 x 64 image 1: Maxpool Layer: 224 x 224 x 64 image, 2 size, 2 stride 2: Convolutional Layer: 112 x 112 x 64 image, 192 filters -> 112 x 112 x 192 image 3: Maxpool Layer: 112 x 112 x 192 image, 2 size, 2 stride 4: Convolutional Layer: 56 x 56 x 192 image, 128 filters -> 56 x 56 x 128 image 5: Convolutional Layer: 56 x 56 x 128 image, 256 filters -> 56 x 56 x 256 image 6: Convolutional Layer: 56 x 56 x 256 image, 256 filters -> 56 x 56 x 256 image 7: Convolutional Layer: 56 x 56 x 256 image, 512 filters -> 56 x 56 x 512 image 8: Maxpool Layer: 56 x 56 x 512 image, 2 size, 2 stride 9: Convolutional Layer: 28 x 28 x 512 image, 256 filters -> 28 x 28 x 256 image 10: Convolutional Layer: 28 x 28 x 256 image, 512 filters -> 28 x 28 x 512 image 11: Convolutional Layer: 28 x 28 x 512 image, 256 filters -> 28 x 28 x 256 image 12: Convolutional Layer: 28 x 28 x 256 image, 512 filters -> 28 x 28 x 512 image 13: Convolutional Layer: 28 x 28 x 512 image, 256 filters -> 28 x 28 x 256 image 14: Convolutional Layer: 28 x 28 x 256 image, 512 filters -> 28 x 28 x 512 image 15: Convolutional Layer: 28 x 28 x 512 image, 256 filters -> 28 x 28 x 256 image 16: Convolutional Layer: 28 x 28 x 256 image, 512 filters -> 28 x 28 x 512 image 17: Convolutional Layer: 28 x 28 x 512 image, 512 filters -> 28 x 28 x 512 image 18: Convolutional Layer: 28 x 28 x 512 image, 1024 filters -> 28 x 28 x 1024 image 19: Maxpool Layer: 28 x 28 x 1024 image, 2 size, 2 stride 20: Convolutional Layer: 14 x 14 x 1024 image, 512 filters -> 14 x 14 x 512 image 21: Convolutional Layer: 14 x 14 x 512 image, 1024 filters -> 14 x 14 x 1024 image 22: Convolutional Layer: 14 x 14 x 1024 image, 512 filters -> 14 x 14 x 512 image 23: Convolutional Layer: 14 x 14 x 512 image, 1024 filters -> 14 x 14 x 1024 image 24: Convolutional Layer: 14 x 14 x 1024 image, 1024 filters -> 14 x 14 x 1024 image 25: Convolutional Layer: 14 x 14 x 1024 image, 1024 filters -> 7 x 7 x 1024 image 26: Convolutional Layer: 7 x 7 x 1024 image, 1024 filters -> 7 x 7 x 1024 image 27: Convolutional Layer: 7 x 7 x 1024 image, 1024 filters -> 7 x 7 x 1024 image 28: Segmentation fault (core dumped)

[AlexanderFreytag]

Hi,

sorry for the late reply.

In fact, I did never worked with the python scripts for generating/converting the data. If you still have troubles with them, I would suggest you to contact Guanghan Ning who originally added these scripts.

Note further that if you are using my mirror from https://github.com/cvjena/darknet there is no need any longer to adapt the labels in src/yolo.c . Instead, I modified the code such that you can directly pass a file with a list of labels in it. No code adaptation any longer :)

Best, Alex

Dipl.-Inf. Alexander Freytag tel: +49 3641 9 46428 fax: +49 3641 9 46372

Research Coordinator Michael Stifel Center Jena for Data-driven and Simulation Science http://www.mscj.uni-jena.de/ Ernst-Abbe-Platz 2 07743 Jena Germany

Member of the Computer Vision Group http://www.inf-cv.uni-jena.de/freytag Institute of Computer Science Friedrich Schiller University Jena Ernst-Abbe-Platz 2 07743 Jena Germany

On Sat, Jul 9, 2016 at 12:55 AM, Darienzhangyy [email protected] wrote:

Hi,I am new to object detection and have been struggling on training Yolo on my data for the past few days. So I would really appreciate any suggestion.

I am able to generate the files like "after_conversion.txt" and "training_list.txt". But I am confused about the path I should create.

Here are my files: ls ~/caffe/darknet/scripts voc_label.py WISCR WISCR_label.py

Here WISCR_label.py is like your convert.py. I am able to run this script but not sure where I should save the output file.

ls WISCR/data annotations Annotations images Images ImageSets labels train.mat

I created my data following the format of PASCAL VOC files, with "Annotations"(a bunch of xml files), "Images"(original JPG) and "ImageSets"(train.txt storing the absolute paths of the images).

"labels" stores the output( ) from the WISCR_label.py.(called train.txt)

I know it is quite important to modify the src/yolo.c file and the path should point to the output file of WISCR_label.py. But what should be in the same parent directory as the "train.txt"? And should I make a "annotations" folder and a "images"(lower-cases) folder in my WISCR/data folder?

Thank you so much.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/cvjena/darknet/issues/2, or mute the thread https://github.com/notifications/unsubscribe/ACm7afguCSDNTjMrQf3f5rpWK8q54gyqks5qTtVcgaJpZM4JIdhL .

[deepkshikha]

Het Darienzhangyy

I am also trying to apply ./darknet yolo train cfg/crater.data cfg/yolo.cfg extraction.conv.weights and getting segmentation fault(core dumped)

If you got any hint please suggest