Error by running YOLOv3 with one class

[phamngocthanhtrung]

Hi everyone,

I already have a Cfg-file and weight file for YOLOv3, which I got with Darknet-Framework after the successful training with one class. I wanted to use these files to analyze the performance of Pytorch. Unfortunately, the running of YOLOv3 failed and I got the following error message:

(env) C:\Users\n01pham\Desktop\Data2Day\pytorch-yolo-v3-master>python detect.py --cfg cfg/yolov3_3l.cfg --weights yolov3_3l.weights --images imgs/Berlin.jpg
Loading network.....
Network successfully loaded
C:\ProgramData\Anaconda3\envs\env\lib\site-packages\torch\nn\modules\upsampling.py:129: UserWarning: nn.Upsample is deprecated. Use nn.functional.interpolate instead.
  warnings.warn("nn.{} is deprecated. Use nn.functional.interpolate instead.".format(self.name))
Traceback (most recent call last):
  File "detect.py", line 242, in <module>
    objs = [classes[int(x[-1])] for x in output if int(x[0]) == im_id]
  File "detect.py", line 242, in <listcomp>
    objs = [classes[int(x[-1])] for x in output if int(x[0]) == im_id]
IndexError: list index out of range

In the detect.py I set the number of classes in 112th line to 1. I also changed the list of class names in the voc.names file to the name of my class. I'd be very happy if you could help me with that.

Best regards

[TiongSun]

The problem is at line 80 of util.py.

If you only have one class, the load_class function at util.py will return an empty list instead of a list with one item. It can be fixed by replacing the load_class function in util.py (line 212) with:

def load_classes(namesfile): fp = open(namesfile, "r") names = [i.strip('\n') for i in fp.readlines()] return names

[kalikhademi]

Thanks @TiongSun for your response it actually resolved this error for me!