jwchoi384/Gaussian_YOLOv3

Issues

• when i train my own dataset, the error "darknet: ./src/parser.c:340: parse_region: Assertion `l.outputs == params.inputs' failed." I have modified my cfg due to the paper.

  #18 opened 5 years ago
  1

• how do you think about NLL loss?

  #17 opened 5 years ago
  6

• Questions about training

  #16 opened 5 years ago
  2

• FP from a false localization can lead fatal accidents?

  #15 opened 5 years ago
  1

• Can I use .weights file to be my pre-trained weights?

  #14 opened 5 years ago
  1

• How to prepare bdd dataset for training?

  #13 opened 5 years ago
  4

• loss is nan

  #12 opened 5 years ago
  5

• Can you explain the loss function (5) in paper?

  #11 opened 5 years ago
  2

• train myown data?

  #10 opened 5 years ago
  1

• Training failure

  #9 opened 5 years ago
  3

• -nogpu lead to Incorrect detection, can you fix it?

  #8 opened 5 years ago
  2

• Do you have any plans for pytorch or tf implementation?

  #7 opened 5 years ago
  2

• what is your iou when you test the bdd100k？

  #6 opened 5 years ago
  2

• in caffe?pytorch?

  #5 opened 5 years ago
  4

• why I use camera ,the process time is too long

  #4 opened 5 years ago
  1

• txt2json

  #3 opened 5 years ago
  5

• Did you compare Gaussian_YOLOv3 with GIoU?

  #2 opened 5 years ago
  1

• KITIT dataset

  #1 opened 5 years ago
  5