dog-qiuqiu/FastestDet

Issues

• AssertionError: 请指定正确的配置文件路径

  #25 opened 2 years ago by yelimei
  1

• train中，直接修改yaml文件中的nc数为1时，在load param时会报错

  #26 opened 2 years ago by qianguanyu12138
  2

• May choose the function `os.path.splitext` ?

  #44 opened a year ago by Z-MiCTrue
  1

• 请问您的数据集标签格式是什么样的呢，是YOLO格式吗

  #41 opened 2 years ago by texas61
  2

• FatestDet线程设置没用，同样设置YOLO-fastestv2正常

  #45 opened a year ago by Zi-x
  1

• example/ncnn 文件下 ncnn 测试问题

  #14 opened 2 years ago by Rolin-C
  3

• can you offer the network structure diagram?

  #17 opened 2 years ago by superbayes
  2

• 无法保存模型权重

  #38 opened 2 years ago by jinyu-118
  2

• 请问大佬这训练过程是否有出错的地方，一直找不到训练出来的模型

  #40 opened 2 years ago by summm
  4

• train.py导出路径

  #20 opened 2 years ago by tututu11
  1

• FastestDet_NCNN

  #12 opened 2 years ago by pengyang1225
  1

• Is SPP class in custom_layers.py means Spatial Pyramid Pooling?

  #48 opened 8 months ago by sayyidan-i
  0

• 小目标label无法参加训练

  #24 opened 2 years ago by ffffwyy
  3

• ncnn出问题

  #21 opened 2 years ago by tututu11
  2

• Set custom input resolution

  #46 opened a year ago by susanin1970
  2

• README weight_AP05:0.253207_280-epoch.pth Colon Issues

  #47 opened a year ago by hefang-x
  0

• 训练时出现错误

  #43 opened 2 years ago by wxzyqqj
  1

• onnxruntime和ncnn速度对比

  #22 opened 2 years ago by ISCLab-Bistu
  2

• 请问cpu如何训练？

  #42 opened 2 years ago by heisechibang
  0

• 使用OpenCV部署FastestDet，包含C++和Python两种版本的程序

  #5 opened 2 years ago by hpc203
  4

• 能否有预训练模型?

  #9 opened 2 years ago by superbayes
  4

• Are the weights of shufflenetv2.pth from ImageNet training ?

  #30 opened 2 years ago by Hrayo712
  2

• 实际检测准确率

  #33 opened 2 years ago by TryHard-LL
  2

• mAP, Precision, Recall

  #37 opened 2 years ago by chequanghuy
  0

• mAP的问题

  #19 opened 2 years ago by winneryoow1
  2

• 请问有论文吗？希望拜读一下

  #32 opened 2 years ago by feidufuzhou
  0

• Right way to convert coco dataset

  #31 opened 2 years ago by dzk9528
  1

• 请教大佬，我训练的过程，发现，越来越小，IOU从0.5降到0.0x, 为什么呢？烦请赐教！

  #29 opened 2 years ago by cncowboy
  0

• 转成ncnn之后模型精度差好多

  #28 opened 2 years ago by 523997931
  2

• MemoryError

  #27 opened 2 years ago by centurions
  0

• 导出的onnx输出shape是 float32[1,85,22,22],如何转化成想要的 (x0,y0,x1,y1,class)的结果

  #23 opened 2 years ago by iwanglian
  1

• 这个模型怎么部署到Android上呢，有没有示例demo呢

  #18 opened 2 years ago by jeekMic
  1

• Is it a MobileOne project?

  #1 opened 2 years ago by Grabber
  3

• 生成的2个预测框几乎重合

  #16 opened 2 years ago by superbayes
  1

• pth文件命名带了冒号

  #15 opened 2 years ago by caizhuojiang
  0

• multi-GPU training

  #10 opened 2 years ago by songtf525
  1

• ncnn implementation

  #4 opened 2 years ago by soham24
  2

• The require.txt file is missing

  #2 opened 2 years ago by dankernel
  2