I am trying to train my Yolov3 model to detect one calss of object. I use the command sudo ./darknet detector train cfg/coco.data cfg/yolov3_test.cfg darknet53.conv.74
My coco.data is set up as so:
classes= 1
train = /home/philip/Downloads/YOLO/images/train.txt
valid = /home/philip/Downloads/YOLO/images/train.txt
names = /home/philip/darknet/AlexeyAB/darknet/data/coco.names
backup = /home/philip/darknet/AlexeyAB/darknet/backup
eval=coco
(I know my training and valid data sets are the same, I just want to see if the training is actually just completes for now.) WHen I run the commnad on terminal, the terminal spits out:
..... Cannot load image "/home/philip/Downloads/YOLO/images/plastic_bags.png 115,74,178,108,0 223,87,298,114,0 254,47,279,58,0 286,64,298,74,0"
STB Reason: can't fopen
Failed to infer label file name (check image extension is supported): /home/philip/Downloads/YOLO/images/plastic_bags.png 115,74,178,108,0 223,87,298,114,0 254,47,279,58,0 286,64,298,74,0
Can't open label file. (This can be normal only if you use MSCOCO): /home/philip/Downloads/YOLO/images/plastic_bags.png 115,74,178,108,0 223,87,298,114,0 254,47,279,58,0 286,64,298,74,0
......
I'm not sure what the points of error in my setup are. Perhaps its the format of some of my files.
I will go through what each paramter looks like.
This is my train.txt:
/home/philip/Downloads/YOLO/images/plastic_bags.png 115,74,178,108,0 223,87,298,114,0 254,47,279,58,0 286,64,298,74,0
This is my coco.names:
plasticwaste
Perhaps its that my paramters aren't all in the same directory? I honestly have no clue. I've tried using sudo on the command because maybe darknet hard time getting permissions to access some files but that doesnt work. Any advice would greatly appreciated.
Why are you using sudo?
Why are you using STB instead of OpenCV?
Why does it look like you have coordinates inside your train.txt file?
It doesn't look like you are setting up the project files correctly. I strongly suggest you read the Darknet FAQ: https://www.ccoderun.ca/programming/darknet_faq/
In addition, I would strongly suggest you use DarkMark to setup the Darknet project correctly. You can find DarkMark here: https://github.com/stephanecharette/DarkMark#what-is-darkmark