biubug6/Face-Detector-1MB-with-landmark

Issues

• 大佬搞过MNN的retinaface部署么，求教啊~

  #35 opened 5 years ago by luqiang6q
  1

• C++_inference _ncnn: How to set opencv path in CmakeList.txt if opencv was installed in Conda's virtual environment?

  #55 opened 3 years ago by jamesjianwang
  1

• 怎么显示关键点

  #62 opened 2 years ago by wzc201391
  0

• Landmark wrong with rotate face 90 degree

  #61 opened 2 years ago by chuongnvk54
  0

• onnx2ncnn的问题

  #31 opened 5 years ago by Vivianyzw
  7

• pytorch模型精度与ncnn模型精度不一致

  #60 opened 3 years ago by 408550969
  2

• collect2: error: ld returned 1 exit status

  #52 opened 4 years ago by chaotianjiao
  3

• Is it possible to have pose estimation?

  #29 opened 5 years ago by abdalazizrashid
  1

• 关于使用四个关键点训练模型的问题？

  #59 opened 3 years ago by Hezhexi2002
  0

• 不同版本性能比较

  #58 opened 3 years ago by JH-Lam
  0

• do I have to reAnnotation the box and landmark coordinaties If change the config image_size

  #42 opened 4 years ago by vgg4resnet
  1

• 如何在样本中增加负样本？

  #14 opened 5 years ago by lyp-deeplearning
  3

• 你好，有关于5个关键点的NME测试标记（widerface）和代码吗？

  #33 opened 5 years ago by Wo-Ki
  2

• Help！！！open failed face.param read_proto_from_binary failed

  #56 opened 3 years ago by Maping1026
  0

• 关于检测其他物体的问题

  #57 opened 3 years ago by 205418367
  0

• How did you get "min_sizes" and "variance" in config.py?

  #36 opened 5 years ago by cannguyen275
  1

• C++_inference _ncnn在MacOS下有问题，problem in MacOS

  #54 opened 3 years ago by jamesjianwang
  0

• 数据集获取

  #43 opened 4 years ago by xiaoguitoucode
  2

• About scale when detect new image

  #53 opened 4 years ago by docongminh
  0

• resize时长宽比的问题

  #51 opened 4 years ago by LeeBC2298
  0

• 为什么Face-Detector-1MB-with-landmark的模型会比原版的好呢？

  #18 opened 5 years ago by tfygg
  3

• Tensorrt Conversion

  #50 opened 4 years ago by prnvjb
  0

• read_proto_from_binary failed

  #49 opened 4 years ago by MLQUE
  0

• image -= (104, 117, 123) 这一步需要消耗时间与图片尺寸成正比

  #48 opened 4 years ago by LUOQING1994
  0

• 为什么ncnn减均值需要6个ms左右

  #46 opened 4 years ago by isyanan1024
  0

• 关于多分类

  #32 opened 5 years ago by duorouputao
  1

• CUDA out of memory when I try to detect face in CelebA dataset.

  #47 opened 4 years ago by khanhnd0408
  1

• 关于转onnx和ncnn的param参数输出不同的问题

  #45 opened 4 years ago by ylc2580
  1

• NCNN library for android

  #19 opened 5 years ago by sanrahul
  1

• run demo on jetson nano

  #21 opened 5 years ago by zhengzhe97
  2

• comment and code not matched in random crop

  #44 opened 4 years ago by SungjunKim1
  0

• onnx模型验证

  #15 opened 4 years ago by yjhuasheng
  3

• About WiderFace test setting

  #40 opened 4 years ago by tabsun
  0

• 特征点增加至68点训练，效果很不好

  #11 opened 5 years ago by github-luffy
  2

• 请问大佬用的什么版本的ncnn?

  #39 opened 4 years ago by whutly123
  0

• 使用原版模型已经转换成功，在使用编译程序的时候报错

  #37 opened 5 years ago by hufengshuo07
  2

• 您好，感谢你提供的代码，请可以在提供一份mobilenetX0.25的在image上面的预训练模型吗

  #30 opened 5 years ago by LLLLLLoki
  2

• 沒复现作者的结果，能给点建议吗？

  #34 opened 5 years ago by goldlee
  0

• 如果只想检测图像中稍大的脸，比如320x180中20x20的脸，有什么办法吗？

  #22 opened 5 years ago by liqiang311
  3

• how to annotate landmark?

  #28 opened 5 years ago by shaqing
  0

• RBF_Final.pth第一个batchnorm的 mean & val 为什么这么大

  #26 opened 5 years ago by WendShadow
  0

• How to design the anchor?

  #27 opened 5 years ago by AnthonyF333
  0

• 请问下 项目下最多人自拍照那张结果用什么模型测试的啊

  #25 opened 5 years ago by DuckJ
  0

• 请问ssh模块去掉了Relu,是怎么考虑的呢？

  #24 opened 5 years ago by deep-practice
  0

• Dense Regression loss

  #23 opened 5 years ago by JitindraFartiyal
  0

• label为-1 仅仅是因为一些人脸无法标注关键点信息, 只是用来回归位置信息(不回归关键点信息)

  #20 opened 5 years ago by yjhuasheng
  0

• 关于损失函数部分confidence的损失值的计算

  #17 opened 5 years ago by HuColin
  1

• How to handle the faces without landmarks?

  #16 opened 5 years ago by xperzy
  0

• MNN Implementation

  #12 opened 5 years ago by PhenomenalOnee
  1

• widerface亲测疑问

  #13 opened 5 years ago by ufoym
  2