apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-14 12:55:06 +08:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

[CVCUDA] Add CV-CUDA support in PaddleSeg (#1761)

Browse Source 
* add cvcuda support in ppseg

* python and pybind

* add resize op, remove concat,std::move

* define resize op
This commit is contained in:
guxukai
2023-04-09 10:38:18 +08:00
committed by GitHub
parent c90aa7bd6f
commit ed19c759df
5 changed files with 128 additions and 96 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
fastdeploy/vision/segmentation/ppseg/model.cc
View File

@@ -79,7 +79,8 @@ bool PaddleSegModel::BatchPredict(const std::vector<cv::Mat>& imgs,
std::vector<FDMat> fd_images = WrapMat(imgs);
// Record the shape of input images
std::map<std::string, std::vector<std::array<int, 2>>> imgs_info;
if (!preprocessor_.Run(&fd_images, &reused_input_tensors_, &imgs_info)) {
preprocessor_.SetImgsInfo(&imgs_info);
if (!preprocessor_.Run(&fd_images, &reused_input_tensors_)) {
FDERROR << "Failed to preprocess input data while using model:"
<< ModelName() << "." << std::endl;
return false;