GitLab

* Add Deformable Convolution operation.

This adds the deformable convolution operation, as described in Deformable Convolutional Networks (https://arxiv.org/abs/1703.06211).

- The code is based on https://github.com/open-mmlab/mmdetection/blob/master/mmdet/ops/dcn/src/deform_conv_cuda.cpp ; the whole code was modified and refactored to remove redundancies and increase clarity, and to adapt it to torchvision.

- The CPU part is a direct copy of the CUDA code; it might make sense to do follow-up adjustments in the CPU code to simplify it / optimize it, or to reuse functionality between CPU and CUDA..

- We also add tests (with a non-trivial set of parameters); they can be made more robust by randomizing the parameters and executing multiple times.

* Update DeformConv to be more consistent w/ Conv2d

* rename some variables and arguments to match Conv2d;
* add optional bias;
* add weight, offset and bias as module parameters;
* remove the n_parallel_imgs parameter;
* Fix __repr__;
* etc..

Initialization of weight and bias is the same as in Conv2d, and
initialization of offsets to zero is the same as in the paper.

This also includes some other small unrelated fixes/improvements.

* Apply clang-format in DeformConv files.

* Import Optional type annotation

* Remove offset param from DeformConv2d module

- We pass the offset in the forward of DeformConv2d, instead of having
an internal parameter. This adds some complexity to creating the module
(e.g. now you have to worry about the output size, to create the
offset), but it gives more flexibility.
- We also use make_tuple for tuple creation, in an attempt to fix error
w/ older compilers.

* Replace abs by std::abs

Old gcc versions were giving wrong results here, because they would
resolve abs as int -> int, thus causing undesired truncation. Replacing
abs by std::abs should allow for correct overloading of abs as float -> float.

* Reorder declarations for clarity

* Reorder weight and offset args in deform_conv2d

We place offset arg before the weight arg, to be more
consistent with DeformConv2d.forward(input, offset)

* Replace abs by std::abs in DeformConv_cuda

* almost working...

* respond to comments

* add empty tensor op, handle different output types in generalized rcnn

* clean ups

* address comments

* more changes

* it's working!

* torchscript bugs

* add script/ eager test

* eval script model

* fix flake

* division import

* py2 compat

* update test, fix arange bug

* import division statement

* fix linter

* fixes

* changes needed for JIT master

* cleanups

* remove imagelist_to

* requested changes

* Make FPN backwards-compatible and torchscript compatible

We remove support for feature channels=0, but support for it was already a bit limited

* Fix ONNX regression

Rename with_bias() to bias(), and output_channels() to out_channels() in C++ conv layer options usage (#1576)

* Fix inconsistent NMS implementation

* Improve tests for NMS

* Remove unnecessary using statement

* added PSRoiAlign and PSRoiPool with C++ autograd and torch ops

* fixed linter errors

* fixed linter errors 2

* fixed linter errors 3

* extend video reader to support fast video probing

* fix c++ lint

* small fix

* allow to accept input video of type torch.Tensor

Revert "Change all torch::nn::init::Nonlinearity::{name} and torch::nn::init::FanMode::{name} to torch::k{name} (#1394)" (#1428)

This reverts commit 8c3cea7f.

Change all torch::nn::init::Nonlinearity::{name} and torch::nn::init::FanMode::{name} to torch::k{name} (#1394)

* Change all torch::nn::init::Nonlinearity::{name} and torch::nn::init::FanMode::{name} to torch::k{name}

* empty commit

* fix lint

* fix lint

* fix lint

Revert "Change all torch::nn::init::Nonlinearity::{name} and torch::nn::init::FanMode::{name} usage to torch::k{name}"

This reverts commit e22c105c.

Change all torch::nn::init::Nonlinearity::{name} and torch::nn::init::FanMode::{name} usage to torch::k{name}

* [video reader] inception commit

* add method save_metadata to class VideoClips in video_utils.py

* add load_metadata() method to VideoClips class

* add Exception to not catch unexpected events such as memory erros, interrupt

* fix bugs in video_plus.py

* [video reader]remove logging. update setup.py

* remove time measurement in test_video_reader.py

* Remove glog and try making ffmpeg finding more robust

* Add ffmpeg to conda build

* Add ffmpeg to conda build [again]

* Make library path finding more robust

* Missing import

* One more missing fix for import

* Py2 compatibility and change package to av to avoid version conflict with ffmpeg

* Fix for python2

* [video reader] support to decode one stream only (e.g. video/audio stream)

* remove argument _precomputed_metadata_filepath

* remove save_metadata method

* add get_metadata method

* expose _precomputed_metadata and frame_rate arguments in video dataset __init__ method

* remove ssize_t

* remove size_t to pass CI check on Windows

* add PyInit__video_reader function to pass CI check on Windows

* minor fix to define PyInit_video_reader symbol

* Make c++ video reader optional

* Temporarily revert changes to test_io

* Revert changes to python files

* Rename files to make it private

* Fix python lint

* Fix C++ lint

* add a functor object EnumClassHash to make Enum class instances usable as key type of std::unordered_map

* fix cpp format check

* Remove C++ extensions in favor of custom ops

* Remove unused custom_ops.cpp file

* Rename _custom_ops.py

* Reorganize functions

* Minor improvements and fixes

* Fix lint

* Fully scriptable ops

* Import types used by annotations

* Make custom ops differentiable

and replace autograd.Function. Use ops unconditionally.

We may consider removing the extension functions in a follow-up.

The code-path is tested by the exisitng tests for differentiability.

* add scripting gradchecks tests and use intlist

* fix implicit tuple conversion for gcc-5

* fix merge

* Revert "Revert "Register Torchvision Ops as Cutom Ops (#1267)" (#1316)"

This reverts commit fe234fc8.

* Make import of C++ extensions lazy

* define python initialization functions for extension

* Fix lint

This reverts commit 78f169b5.

* Register torchvision ops

* install ORT only with python 3

* remane lib + address other comments

* fix lint

* fix lib copy

* find file with pattern instead of suffix

* use relative path

* revert rename and use imp to find lib

* fix typo

* Use Tensor.data_ptr instead of .data

* use pytorch-nightly in CI

* Replaced asserts with TORCH_CHECK

* Fixed an error

* Added mnasnet

* Fixed some stuff

* Fixed some stuff

* Finished MNASNet

* Fixed format error

* Fixed format error

In-place operators should only have method variant. Update here in
preparation of future PyTorch change. For example see
https://github.com/pytorch/pytorch/issues/22707

* Enabling exporting symbols on windows

Small fix to allow for the built library to be used in windows #728

* added macro to allow for exported symbols on windows

* added macro to allow for exported symbols on windows

* removed cmake command

* added dllimport using torchvision_EXPORTS preprocessor

* Added the existing code

* Added squeezenet and fixed some stuff in the other models

* Wrote DenseNet and a part of InceptionV3

Going to clean and check all of the models and finish inception

* Fixed some errors in the models

Next step is writing inception and comparing with python code again.

* Completed inception and changed models directory

* Fixed and wrote some stuff

* fixed maxpoool2d and avgpool2d and adaptiveavgpool2d

* Fixed a few stuff

Moved cmakelists to root and changed the namespace to vision and wrote weight initialization in inception

* Added models namespace and changed cmakelists

the project is now installable

* Removed some comments

* Changed style to pytorch style, added some comments and fixed some minor errors

* Removed truncated normal init

* Changed classes to structs and fixed a few errors

* Replaced modelsimpl structs with functional wherever possible

* Changed adaptive average pool from struct to function

* Wrote a max_pool2d wrapper and added some comments

* Replaced xavier init with kaiming init

* Fixed an error in kaiming inits

* Added model conversion and tests

* Fixed a typo in alexnet and removed tests from cmake

* Made an extension of tests and added module names to Densenet

* Added python tests

* Added MobileNet and GoogLeNet models

* Added tests and conversions for new models and fixed a few errors

* Updated Alexnet ad VGG

* Updated Densenet, Squeezenet and Inception

* Added ResNexts and their conversions

* Added tests for ResNexts

* Wrote tools nessesary to write ShuffleNet

* Added ShuffleNetV2

* Fixed some errors in ShuffleNetV2

* Added conversions for shufflenetv2

* Fixed the errors in test_models.cpp

* Updated setup.py

* Fixed flake8 error on test_cpp_models.py

* Changed view to reshape in forward of ResNet

* Updated ShuffleNetV2

* Split extensions to tests and ops

* Fixed test extension

* Fixed image path in test_cpp_models.py

* Fixed image path in test_cpp_models.py

* Fixed a few things in test_cpp_models.py

* Put the test models in evaluation mode

* Fixed registering error in GoogLeNet

* Updated setup.py

* write test_cpp_models.py with unittest

* Fixed a problem with pytest in test_cpp_models.py

* Fixed a lint problem

* #944
  MSBuild Compile time casting Error

* #944
 MSBuild Error  static_cast<Long> to static_cast<int64_t>

* Add eval.py
 Not Work find_contours

* Remove unnecessary file

* Lint

Updated nms_cuda signature to accept detections and scores as separate tensors.
This also required updating the indexing in the NMS CUDA kernel.

Also made the iou_threshold parameter name consistent across implementations.

* Initial layout for layers with cpp extensions

* Move files around

* Fix import after move

* Add support for multiple types to ROIAlign

* Different organization

CUDA extensions work now

* Cleanups

* Reduce memory requirements for backwards

* Replace runtime_error by AT_ERROR

* Add nms test

* Add support for compilation using CPP extensions

* Change folder structure

* Add ROIPool cuda

* Cleanups

* Add roi_pool.py

* Fix lint

* Add initial structures folder for bounding boxes

* Assertion macros compatible with pytorch master (#540)

* Support for ROI Pooling (#592)

* ROI Pooling with tests. Fix for cuda context in ROI Align.

* renamed bottom and top to follow torch conventions

* remove .type().tensor() calls in favor of the new approach to tensor initialization (#626)

* Consistent naming for rois variable (#627)

* remove .type().tensor() calls in favor of the new approach to tensor initialization

* Consistent naming for rois variable in ROIPool

* ROIPool: Support for all datatypes (#632)

* Use of torch7 naming scheme for ROIAlign forward and backward

* use common cuda helpers in ROIAlign

* use .options() in favor of .type() where applicable

* Added tests for forward pass of ROIAlign, as well as more consistent naming scheme for CPU vs CUDA

* working ROIAlign cuda backwards pass

* working ROIAlign backwards pass for CPU

* added relevant headers for ROIAlign backwards

* tests for ROIAlign layer

* replace .type() with .options() for tensor initialization in ROIAlign layers

* support for Half types in ROIAlign

* gradcheck tests for ROIAlign

* updated ROIPool on CPU to work with all datatypes

* updated and cleaned tests for ROI Pooling

* Fix rebase problem

* Remove structures folder

* Improve cleanup and bugfix in test_layers

* Update C++ headers

* Add CUDAGuard to cu files

* Add more checks to layers

* Add CUDA NMS and tests

* Add multi-type support for NMS CUDA

* Avoid using THCudaMalloc

* Add clang-format and reformat c++ code

* Remove THC includes

* Rename layers to ops

* Add documentation and rename functions

* Improve the documentation a bit

* Fix some lint errors

* Fix remaining lint inssues

* Area computation doesn't add +1 in NMS

* Update CI to use PyTorch nightly

* Make NMS return indices sorted according to the score

* Address reviewer comments

* Lint fixes

* Improve doc for roi_align and roi_pool

* move to xenial

* Fix bug pointed by @lopuhin

* Fix RoIPool reference implementation in Python 2

Also fixes a bug in the clip_boxes_to_image -- this function needs a test!

* Remove change in .travis