mendelxu/ann

semantic segmentation，pytorch，non-local

ann's Issues

Losses

Your losses.py at https://github.com/MendelXu/ANN/blob/master/models/seg/loss/seg_modules.py has a number of different loss functions including some for embedding layers which are not mentioned in your paper. Could you kindly provide some explanation for what these losses mean? Would love to learn more about them, they seem quite interesting. Thanks!

sim_map = (self.key_channel**-.5) * sim_map

sim_map = (self.key_channel**-.5) * sim_map
你好，请问这句代码到底有什么作用啊,他的前面不是已经有两个相乘了嘛，为什么还有这句代码啊
sim_map= torch.matmul(query, key)

Can this only be trained on multiple Gpus? Can you train on a single GPU

@MendelXu Can this only be trained on multiple Gpus? Can you train on a single GPU？thanks

Feature visualization

Hi,

Thank you for this wonderful work,

I was wandering about the feature visualization in the sup. mat., can you please describe how such visualization were computed (or point me to some github repo to generate such vizualizations).

Thank you.

Can this model be used for human segmentation？

@MendelXu thanks for this great work, Can this model be used for human segmentation？for example :hair ,face and so on

thanks

put the downloaded trained [model] to where?

-Validation
Download trained model to
checkpoints/seg/cityscapes/fs_annn_cityscapes_segohem_latest.pth and test on validation set with single scale.

bash scripts/seg/cityscapes/run_fs_annn_cityscapes_seg.sh val ohem

Hello, would you like to put the downloaded trained [model] in the cityscapes directory or the .pth directory, or change the name of the model to fs_annn_cityscapes_segohem_latest.pth?I changed the name of the downloaded model and put it in the cityscapes directory, and the following error occurred.

2022-03-27 18:54:21,362 INFO    [controller.py, 67] Testing start...
Traceback (most recent call last):
  File "main.py", line 201, in <module>
    Controller.test(runner)
  File "/peng/zqk/ANN-master/methods/tools/controller.py", line 84, in test
    runner.test(test_dir, out_dir)
  File "/peng/zqk/ANN-master/methods/seg/fcn_segmentor_test.py", line 48, in test
    total_logits = self.ss_test(data_dict)
  File "/peng/zqk/ANN-master/methods/seg/fcn_segmentor_test.py", line 88, in ss_test
    results = self._predict(data_dict)
  File "/peng/zqk/ANN-master/methods/seg/fcn_segmentor_test.py", line 213, in _predict
    results = self.seg_net.forward(data_dict['img'])
  File "/home/usr2/anaconda2/envs/zqk36/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 143, in forward
    outputs = self.parallel_apply(replicas, inputs, kwargs)
  File "/home/usr2/anaconda2/envs/zqk36/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 153, in parallel_apply
    return parallel_apply(replicas, inputs, kwargs, self.device_ids[:len(replicas)])
  File "/home/usr2/anaconda2/envs/zqk36/lib/python3.6/site-packages/torch/nn/parallel/parallel_apply.py", line 36, in parallel_apply
    assert len(modules) == len(inputs)
AssertionError
evaluate the result...
ERROR: Found no prediction for ground truth /data/DataSet/Cityscapes/val/label/lindau_000041_000019_leftImg8bit.png

tool for computation and memory statistics

First of all thank you for this wonderful work, and for providing the implementation.

Feel free to close this issue if this is not the place to ask such questions. I really liked your computation and memory statistics, I was wandering what tool you've used to get such statistics.

Thank you very much

精度

你好作者请问你实现了这个吗，我实现了ccnet danet ocnet annn这几个基于non-local原理的模型，发现结果反而更差，我不知道为什么，请问是否和torch的版本有关，或者是训练的图片长宽必须是偶数还是奇数有关系啊？？

subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.

python 3.6.4
cuda 9.0
gcc 5.5.0
I meet problem when I run bash scripts/seg/cityscapes/run_fs_annn_cityscapes_seg.sh train tag
the first error is subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.
can you solve the problem?

Traceback (most recent call last): File "main.py", line 182, in <module> runner = method_selector.select_seg_method() File "/data/lumeng/ANN/methods/method_selector.py", line 104, in select_seg_method return SEG_METHOD_DICT[key](self.configer) File "/data/lumeng/ANN/methods/seg/fcn_segmentor.py", line 44, in __init__ self._init_model() File "/data/lumeng/ANN/methods/seg/fcn_segmentor.py", line 47, in _init_model self.seg_net = self.seg_model_manager.semantic_segmentor() File "/data/lumeng/ANN/models/seg/model_manager.py", line 38, in semantic_segmentor model = SEG_MODEL_DICT[model_name](self.configer) File "/data/lumeng/ANN/models/seg/nets/annn.py", line 17, in __init__ self.backbone = BackboneSelector(configer).get_backbone() File "/data/lumeng/ANN/models/backbones/backbone_selector.py", line 31, in get_backbone model = ResNetBackbone(self.configer)(**params) File "/data/lumeng/ANN/models/backbones/resnet/resnet_backbone.py", line 176, in __call__ orig_resnet = self.resnet_models.deepbase_resnet101() File "/data/lumeng/ANN/models/backbones/resnet/resnet_models.py", line 256, in deepbase_resnet101 norm_type=self.configer.get('network', 'norm_type'), **kwargs) File "/data/lumeng/ANN/models/backbones/resnet/resnet_models.py", line 107, in __init__ ('bn1', ModuleHelper.BatchNorm2d(norm_type=norm_type)(64)), File "/data/lumeng/ANN/models/tools/module_helper.py", line 89, in BatchNorm2d from encoding.nn import BatchNorm2d File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/__init__.py", line 13, in <module> from . import nn, functions, parallel, utils, models, datasets, transforms File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/nn/__init__.py", line 12, in <module> from .encoding import * File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/nn/encoding.py", line 18, in <module> from ..functions import scaled_l2, aggregate, pairwise_cosine File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/functions/__init__.py", line 2, in <module> from .encoding import * File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/functions/encoding.py", line 14, in <module> from .. import lib File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/lib/__init__.py", line 27, in <module> build_directory=gpu_path, verbose=False) File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 645, in load is_python_module) File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 814, in _jit_compile with_cuda=with_cuda) File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 863, in _write_ninja_file_and_build _build_extension_module(name, build_directory, verbose) File "/home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 959, in _build_extension_module raise RuntimeError(message) RuntimeError: Error building extension 'enclib_gpu': [1/7] /usr/local/cuda/bin/nvcc -DTORCH_EXTENSION_NAME=enclib_gpu -DTORCH_API_INCLUDE_EXTENSION_H -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include/torch/csrc/api/include -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include/TH -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include/THC -isystem /usr/local/cuda/include -isystem /home/lumeng/anaconda3/envs/pytorch3/include/python3.6m -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --compiler-options '-fPIC' --expt-extended-lambda -std=c++11 -c /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/lib/gpu/nms_kernel.cu -o nms_kernel.cuda.o FAILED: nms_kernel.cuda.o /usr/local/cuda/bin/nvcc -DTORCH_EXTENSION_NAME=enclib_gpu -DTORCH_API_INCLUDE_EXTENSION_H -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include/torch/csrc/api/include -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include/TH -isystem /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/torch/lib/include/THC -isystem /usr/local/cuda/include -isystem /home/lumeng/anaconda3/envs/pytorch3/include/python3.6m -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --compiler-options '-fPIC' --expt-extended-lambda -std=c++11 -c /home/lumeng/anaconda3/envs/pytorch3/lib/python3.6/site-packages/encoding/lib/gpu/nms_kernel.cu -o nms_kernel.cuda.o

which line is the Pyramid Pooling?

which line is the Pyramid Pooling?

subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.

See https://github.com/pytorch/pytorch/blob/master/CONTRIBUTING.md for help
with compiling PyTorch from source.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

                          !! WARNING !!

platform=sys.platform))
Traceback (most recent call last):
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 946, in _build_extension_module
check=True)
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/subprocess.py", line 438, in run
output=stdout, stderr=stderr)
subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "main.py", line 182, in
runner = method_selector.select_seg_method()
File "/home/docker/data/licong/ANN/methods/method_selector.py", line 104, in select_seg_method
return SEG_METHOD_DICTkey
File "/home/docker/data/licong/ANN/methods/seg/fcn_segmentor.py", line 44, in init
self.init_model()
File "/home/docker/data/licong/ANN/methods/seg/fcn_segmentor.py", line 47, in init_model
self.seg_net = self.seg_model_manager.semantic_segmentor()
File "/home/docker/data/licong/ANN/models/seg/model_manager.py", line 38, in semantic_segmentor
model = SEG_MODEL_DICTmodel_name
File "/home/docker/data/licong/ANN/models/seg/nets/annn.py", line 17, in init
self.backbone = BackboneSelector(configer).get_backbone()
File "/home/docker/data/licong/ANN/models/backbones/backbone_selector.py", line 31, in get_backbone
model = ResNetBackbone(self.configer)(**params)
File "/home/docker/data/licong/ANN/models/backbones/resnet/resnet_backbone.py", line 176, in call
orig_resnet = self.resnet_models.deepbase_resnet101()
File "/home/docker/data/licong/ANN/models/backbones/resnet/resnet_models.py", line 256, in deepbase_resnet101
norm_type=self.configer.get('network', 'norm_type'), **kwargs)
File "/home/docker/data/licong/ANN/models/backbones/resnet/resnet_models.py", line 107, in init
('bn1', ModuleHelper.BatchNorm2d(norm_type=norm_type)(64)),
File "/home/docker/data/licong/ANN/models/tools/module_helper.py", line 89, in BatchNorm2d
from encoding.nn import BatchNorm2d
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/init.py", line 13, in
from . import nn, functions, parallel, utils, models, datasets, transforms
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/nn/init.py", line 12, in
from .encoding import *
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/nn/encoding.py", line 18, in
from ..functions import scaled_l2, aggregate, pairwise_cosine
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/functions/init.py", line 2, in
from .encoding import *
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/functions/encoding.py", line 14, in
from .. import lib
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/lib/init.py", line 27, in
build_directory=gpu_path, verbose=False)
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 645, in load
is_python_module)
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 814, in jit_compile
with_cuda=with_cuda)
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 863, in write_ninja_file_and_build
build_extension_module(name, build_directory, verbose)
File "/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 959, in build_extension_module
raise RuntimeError(message)
RuntimeError: Error building extension 'enclib_gpu': [1/7] /usr/local/cuda/bin/nvcc -DTORCH_EXTENSION_NAME=enclib_gpu -DTORCH_API_INCLUDE_EXTENSION_H -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/torch/csrc/api/include -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/TH -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/THC -isystem /usr/local/cuda/include -isystem /data/yicong.lc/anaconda2/envs/py36_ann/include/python3.6m -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS -D__CUDA_NO_HALF_CONVERSIONS -D__CUDA_NO_HALF2_OPERATORS --compiler-options '-fPIC' --expt-extended-lambda -std=c++11 -c /home/docker/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/lib/gpu/nms_kernel.cu -o nms_kernel.cuda.o
FAILED: nms_kernel.cuda.o
/usr/local/cuda/bin/nvcc -DTORCH_EXTENSION_NAME=enclib_gpu -DTORCH_API_INCLUDE_EXTENSION_H -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/torch/csrc/api/include -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/TH -isystem /data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/THC -isystem /usr/local/cuda/include -isystem /data/yicong.lc/anaconda2/envs/py36_ann/include/python3.6m -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --compiler-options '-fPIC' --expt-extended-lambda -std=c++11 -c /home/docker/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/encoding/lib/gpu/nms_kernel.cu -o nms_kernel.cuda.o
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/pytypes.h:1165:206: error: expansion pattern ‘pybind11::detail::negation<std::is_same<pybind11::detail::bools<pybind11::detail::negation<std::is_base_of<pybind11::arg, Args> >::value ..., pybind11::detail::negation<std::is_same<pybind11::detail::kwargs_proxy, Args> >::value ..., true>, pybind11::detail::bools<true, pybind11::detail::negation<std::is_base_of<pybind11::arg, Args> >::value ..., pybind11::detail::negation<std::is_same<pybind11::detail::kwargs_proxy, Args> >::value ...> > >::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/pytypes.h:1165:215: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/pytypes.h:1165:392: error: expansion pattern ‘pybind11::detail::negation<std::is_same<pybind11::detail::bools<pybind11::detail::negation<std::is_base_of<pybind11::arg, Args> >::value ..., pybind11::detail::negation<std::is_same<pybind11::detail::kwargs_proxy, Args> >::value ..., true>, pybind11::detail::bools<true, pybind11::detail::negation<std::is_base_of<pybind11::arg, Args> >::value ..., pybind11::detail::negation<std::is_same<pybind11::detail::kwargs_proxy, Args> >::value ...> > >::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/pytypes.h:1165:395: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/pytypes.h:1165:397: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/pytypes.h:1165:397: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/pytypes.h:1165:412: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:149: error: expansion pattern ‘std::is_copy_constructible<_Tp>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:240: error: expansion pattern ‘std::is_same<typename Container::value_type&, typename Container::reference>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:249: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:249: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:312: error: expansion pattern ‘std::is_copy_constructible<_Tp>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:403: error: expansion pattern ‘std::is_same<typename Container::value_type&, typename Container::reference>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:406: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:406: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:408: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:408: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:423: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:766:425: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:126: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:170: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char16_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:214: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char32_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:257: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, wchar_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:260: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:260: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:260: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:260: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:960:262: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:138: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:186: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char16_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:234: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char32_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:281: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, wchar_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:284: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:284: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:284: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:284: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:286: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:301: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1250:303: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:121: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:169: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char16_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:217: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, char32_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:264: error: expansion pattern ‘pybind11::detail::negation<std::is_same<CharT, wchar_t> >’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:267: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:267: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:267: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:267: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:269: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:284: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1261:286: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:112: error: expansion pattern ‘std::is_void<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:137: error: expansion pattern ‘std::is_pointer<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:164: error: expansion pattern ‘std::is_reference<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:187: error: expansion pattern ‘std::is_const<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:190: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:190: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:190: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:190: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:192: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:201: error: expected parameter pack before ‘...’
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:251: error: expansion pattern ‘pybind11::detail::negation<pybind11::detail::is_copy_constructible >::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:294: error: expansion pattern ‘std::is_move_constructible<_Tp>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:392: error: expansion pattern ‘std::is_same<decltype (declval<pybind11::detail::type_caster<typename pybind11::detail::intrinsic_type::type, void> >().operator T&()), T&>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:401: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:401: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:401: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:401: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:453: error: expansion pattern ‘std::is_void<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:478: error: expansion pattern ‘std::is_pointer<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:505: error: expansion pattern ‘std::is_reference<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:528: error: expansion pattern ‘std::is_const<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:531: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:531: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:531: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:531: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:533: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:542: error: expected parameter pack before ‘...’
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:592: error: expansion pattern ‘pybind11::detail::negation<pybind11::detail::is_copy_constructible >::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:635: error: expansion pattern ‘std::is_move_constructible<_Tp>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:733: error: expansion pattern ‘std::is_same<decltype (declval<pybind11::detail::type_caster<typename pybind11::detail::intrinsic_type::type, void> >().operator T&()), T&>::value’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:736: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:736: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:736: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:736: error: template argument 5 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:738: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:738: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:753: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1583:755: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:121: error: expansion pattern ‘std::is_void<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:146: error: expansion pattern ‘std::is_pointer<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:173: error: expansion pattern ‘std::is_reference<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:196: error: expansion pattern ‘std::is_const<_Tp>’ contains no argument packs
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:199: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:199: error: template argument 2 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:199: error: template argument 3 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:199: error: template argument 4 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:201: error: template argument 1 is invalid
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:210: error: expected parameter pack before ‘...’
/data/yicong.lc/anaconda2/envs/py36_ann/lib/python3.6/site-packages/torch/lib/include/pybind11/cast.h:1590:250: error: expansion pattern ‘pybind11::detail::negation<pybind11::detail::move_always >::value’ contains no argument packs

Same transformation for query and key in APNB?

Hi, I found that the query and key in APNB share the same transformation, see here, which seems to be different from the implementation depicted in Fig.3, where key and value share transformation.
Why? And which is better?
Please correct me if I'm making mistake here. :)

train on my own datasets

@MendelXu sorry for asking questions again,i want to train this model on my own dataset with only two types,what should i do, thanks

how to train my Custom Dataset？

Can you give a tutorial show to train my custom dataset？

关于apnb模块中的 self.f_query = self.f_key？

你好，请问在apnb中此处权重共享，此处设置与nonlocal和afnb中不一样，请问使self.f_query = self.f_key是否会对结果造成影响

i hope to see your code

hello, when can you post your code?

Same transformation for query and key in APNB?

我在#16看到了相同的问题，但是没有解决我的困惑。
“Hi, the implementation of sharing the query and key branch follows Fig. 3 in the arxiv version.”
但是我在arxiv version fig. 3中看到的是key和value共享，而代码实现中是query和key共享，这有什么区别吗？

Question about AFNP & APNB in Fig3

Gook work!
I have some places that I don’t quite understand. AFNB in Sec 3.3 E.q.(14-16) seems there is no key value transormation (θ/φ), its more like APNB in Fig. 3. And the APNB in Sec 3.2 seems to match the AFNB in Fig. 3. Is something that I misunderstand?

When will the source code be released?

Nice paper ! 😆

When will the source code be released?
I'm trying to implement it now, but I don't know the details of implementation.
So I want to refer to your source code.

trained_model on NYUv

could you please release a trained model on NYUv2 dataset? I try to use your network with resnet101-dilated8 to train NYUv2 data,but can only reach 41%miou. Or could please add a trainning scripts on NYUv2.THANKS!

afnb, apnb, does ANN have 1d ?

Hello, thank you very much for your work, nonlocal has 1d, 2d, 3d, I want to know, do your afnb, apnb, ANN have 1d ?