Non-Local Pooling

Update 2.0.1

Now NonLocalPool2d will not use PixelShuffle by default since it may harm the performance. NonLocalPool3d still uses PixelShuffle3d to reduce the token number.

Now you can determine the output channel just like all the other learnable pooling methods (former version force out_channel=in_channel). However, if you specify out_channel, MaxPool would not work then since they cannot be added together. Leave out_channel to None to make the module works like before.

To use NonLocalPooling for your PyTorch project:

pip install nonlocalpooling

from nonlocalpooling.pool import NonLocalPool2d, NonLocalPool3d

Non-Local Pooling can be used to substitue your original PyTorch pooling methods.

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution for nonlocalpooling 2.0.1
File	Size	Uploaded
nonlocalpooling-2.0.1.tar.gz	4.1 kB	May 26, 2022	Details

Table of built distributions (wheels) for nonlocalpooling 2.0.1
File	Interpreter	ABI	Platform	Reset
nonlocalpooling-2.0.1-py3-none-any.whl 4.1 kB May 26, 2022	Python 3	none	any	Details

Total release size: 8.2 kB

Download URL	nonlocalpooling-2.0.1.tar.gz
Size	4.1 kB
Tags	Source
SHA-256 checksum How to use checksums	`ffff4f1e3cf9e3eaae1b08549f81b3d3eecd036812d4f56f861da83410e1f60e`
BLAKE2b-256 checksum How to use checksums	`ae8f2004b588d3988680647c89f1cc45d8d2cdb36f9b08a79d17d419ad0fc1d1`
Upload date	May 26, 2022
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/4.0.0 CPython/3.8.3`

Download URL	nonlocalpooling-2.0.1-py3-none-any.whl
Size	4.1 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`84c9295cd9eee6bab1638bddb63b4e270c6b710574da3c43cd49beb7a840632c`
BLAKE2b-256 checksum How to use checksums	`9ecaf6d21e189d526d03aa4ff451ff30e495757c448a833251d267107ee32b15`
Upload date	May 26, 2022
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/4.0.0 CPython/3.8.3`