Understanding darknet's yolo.cfg config files

后端 未结 4 1289
说谎
说谎 2021-01-30 13:38

I have searched around the internet but found very little information around this, I don\'t understand what each variable/value represents in yolo\'s .cfg files. So

4条回答
  •  慢半拍i
    慢半拍i (楼主)
    2021-01-30 14:39

    batch the number of images chosen in each batch to reduce loss

    subdivisions division of batch size to no. of sub batches for parallel processing

    decay is a learning parameter and as specified in the journal a momentum of 0.9 and decay of 0.0005 is used

    momentum is a learning parameter and as specified in the journal a momentum of 0.9 and decay of 0.0005 is used

    channels Channels refers to the channel size of the input image(3) for a BGR image

    filters the number of filters used for a CNN algorithm

    activation the activation function of CNN: mostly Leaky RELU function is used ( what i have seen mostly in the configuration files)

提交回复
热议问题