名称 最后提交 最后更新
..
details Document transform.h and fix cpplint errors (#9913)
dynload test=develop
CMakeLists.txt fix script issue
assert.h Print assert failure id in lookup_table_op (#14698)
cpu_helper.cc add Set/GetCPUNumThreads api
cpu_helper.h move SetNumThreads to platform
cpu_helper_test.cc move SetNumThreads to platform
cpu_info.cc remove jit namespace
cpu_info.h Remove the useless definition
cpu_info_test.cc Fix cpplint errors with paddle/fluid/platform/cpu_info* (#9708)
cuda_device_function.h "fix float16 ShuffleDownSync Bug" (#12756)
cuda_device_guard.cc Refine code
cuda_device_guard.h Refine code
cuda_helper_test.cu Fix compile error
cuda_primitives.h Fix/float16 style (#12446)
cuda_profiler.h Update documentation (#9918)
cudnn_helper.h fix unit test cases
cudnn_helper_test.cc "fix link error" (#13545)
device_context.cc Fea/fuse conv elementwise add fuse (#14669)
device_context.h merge develop
device_context_test.cu Move fluid::framework::InitDevices into fluid::platform (#11757)
device_tracer.cc Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into my-cool-stuff
device_tracer.h test=develop
enforce.cc Fix the grammar in copyright. (#8403)
enforce.h Fix bug in Windows VC 2010
enforce_test.cc refine by reviewer's advice
float16.h Fix Eigen macro when using GPU
float16_test.cc float16 type support enhance (#12181)
float16_test.cu float16 type support enhance (#12181)
for_range.h Revert "Some trivial optimization (#13530)"
gpu_info.cc [Feature] multi process multi gpu dist training, boost v100 performance by 20% (#14661)
gpu_info.h [Feature] multi process multi gpu dist training, boost v100 performance by 20% (#14661)
hostdevice.h Fix the grammar in copyright. (#8403)
init.cc remove jit namespace
init.h Windows/online (#14474)
init_test.cc Move fluid::framework::InitDevices into fluid::platform (#11757)
lock_guard_ptr.h Clean LockGuardPtr
macros.h Revert "cherry picked windows patches."
mkldnn_helper.h Implement conv3d with mkldnn library (test=develop)
mkldnn_reuse.h - Added reusing MKL-DNN primitives for Transpose MKL-DNN op
nccl_helper.h Change tensor uses proto::VarType::type
ngraph_helper.h Enable mul operator for a ngraph engine (#14801)
place.cc add unit test
place.h Remove place hash
place_test.cc Fix the grammar in copyright. (#8403)
port.h add ctc support for windows
profiler.cc Windows/online (#14474)
profiler.h Windows/online (#14474)
profiler.proto Better timeline
profiler_test.cc Code cleanup in the profiler code. (#9782)
stream_callback_manager.cc
stream_callback_manager.h
transform.h
transform_test.cu
variant.h

项目简介

Object Detection toolkit based on PaddlePaddle. It supports object detection, instance segmentation, multiple object tracking and real-time multi-person keypoint detection.

:rocket: Github 镜像仓库 :rocket:

源项目地址 :arrow_down: :arrow_down: :arrow_down:

https://github.com/PaddlePaddle/PaddleDetection

发行版本

当前项目没有发行版本

贡献者 41

全部贡献者

开发语言

  • Python 92.8 %
  • C++ 3.5 %
  • Jupyter Notebook 1.6 %
  • Cuda 1.1 %
  • CMake 0.7 %