LibtorchSegmentation

LibtorchSegmentation copied to clipboard

关于模型结构

3

你好，看了下unet结构的实施，几个问题请教下： - 分割头上最后为什么还接了一层upsample，按道理conv出来的特征就是(b,n_class,h,w)已经恢复到原图大小了，那这层上采样貌似不是必要的。 - dice_loss的实现：发现dice_coef的公式用了dice_coef=inter/(pred_area + label_area - inter)，但这不是iou的公式吗？dice_coef不是应该等于2*inter/(pred_area + label_area)吗？

ximitiejiang

Libtorch现阶段支持复数运算，博主什么时候能写下复数相关的运算的博客吗，等待更新。

2

看了您写的Libtorch的博客，受益匪浅，慕名而来，star了一下您的项目。 Libtorch现阶段支持复数运算，博主什么时候能写下复数相关的运算的博客吗，等待更新。 博主有没有相关资料的推荐，官网只给了接口，啥也没讲……

majestichou

DiceLoss function was really iou？

1

DiceLoss function in loss.h，just like iou.

julinfn

How to convert pt-model to ONNX-model

1

How to convert pt-model to ONNX-model? I need to load and run an ONNX-model using opencv. Thanks!

Jessicahjj

May you supply a multi-class semantic segmentation example, thanks!

panovr

how can i use Quantization-aware training in this project

steven-spec

there is no main.cpp in the code

1

cannot run the CMakeLists file following the instructions. did not find the src/main.cpp

yulai001

network

2

Author, I have a question for you. I want to change the network to UNet, and change the FPN in the train.cpp and resnet34.cpp files to UNet, but the training...

yuanhs1996

Hi! Can the trained models be converted to onnx runtime? Thanks!

antithing

错误 LNK1104 无法打开文件“segmentation.lib” resnet34 E:\TensorFlow\LibtorchSegmentation-main\test\build\LINK 1

teddyTang9